Growth and Adjustment in East Asia and Latin America

De Gregorio, José; Lee, Jong-wha Working Paper Growth and Adjustment in East Asia and Latin America ADBI Research Paper Series, No. 54 Provided in Cooperation with: Asian Development Bank Institute (ADBI), Tokyo Suggested Citation: De Gregorio, José; Lee, Jong-wha (2004) : Growth and Adjustment in East Asia and Latin America, ADBI Research Paper Series, No. 54, Asian Development Bank Institute (ADBI), Tokyo, https://hdl.handle.net/11540/4159 This Version is available at: https://hdl.handle.net/10419/111144 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/3.0/igo/ ADB INSTITUTE RESEARCH PAPER 54 Growth and Adjustment in East Asia and Latin America José De Gregorio and Jong-Wha Lee February 2004 Using a rigorous econometric approach the authors examine comparative economic performance over the period 1970-2000. Differences in growth performance and macroeconomic adjustment in East Asia and Latin America are found to be largely due to differences in fundamental growth factors such as investment, human resources, fertility, institutional quality, macroeconomic stability and trade openness. Neither differences in initial income nor external shocks to terms of trade are important in explaining lower growth in the Latin American region, refuting the rationale behind the inward looking trade strategy followed by many Latin American countries in the 1960s. Balance-of-payments (BOP) crises have also contributed significantly to differences in growth performance, but the output costs of BOP crises can be reduced by international liquidity, financial soundness, real exchange rate depreciation and sound monetary policy. adbi.org ADB Institute Research Paper Series No. 54 February 2004 Growth and Adjustment in East Asia and Latin America José De Gregorio and Jong-Wha Lee ADB INSTITUTE RESEARCH PAPER 54 Additional copies of the paper are available free from the Asian Development Bank Institute, 8t h Floor, Kasumigaseki Building, 3-2-5 Kasumigaseki, Chiyoda-ku, Tokyo 100-6008, Japan. Attention: Publications. Also online at www.adbi.org Copyright © 2004 Asian Development Bank Institute. All rights reserved. Produced by ADB Publishing. The Research Paper Series primarily disseminates selected work in progress to facilitate an exchange of ideas within the Institute’s constituencies and the wider academic and policy communities. The findings, interpretations, and conclusions are the author’s own and are not necessarily endorsed by the Asian Development Bank Institute. They should not be attributed to the Asian Development Bank, its Boards, or any of its member countries. They are published under the responsibility of the Dean of the ADB Institute. The Institute does not guarantee the accuracy or reasonableness of the contents herein and accepts no responsibility whatsoever for any consequences of its use. The term “country”, as used in the context of the ADB, refers to a member of the ADB and does not imply any view on the part of the Institute as to sovereignty or independent status. Names of countries or economies mentioned in this series are chosen by the authors, in the exercise of their academic freedom, and the Institute is in no way responsible for such usage. II ABSTRACT This paper compares the experience of growth performance and macroeconomic adjustment between East Asia and Latin America from 1970 to 2000. We find that the difference in growth performance between the two regions can be largely attributed to the differences in fundamental growth factors such as investment rate, human resources, fertility, institutional quality, macroeconomic stability and degree of trade openness. We also discuss the role of the quality of education and differences in inequality. Balance-of-payments crisis shocks have also contributed significantly to differences in growth performance. Our analysis reveals that growth rates tend to fall only temporarily following balance-of- payments crises and then rebound to the pre-crisis levels, producing the typical v-type pattern for output in both East Asia and Latin America. However, a balance-of-payments crisis is often associated with a large decline of growth rates and develops into another crisis. Analyzing the determinants of the output cost of crises, we find that international liquidity, financial soundness, real exchange rate depreciation and monetary policy play a critical role in reducing output losses. III TABLE OF CONTENTS Abstract III Table of Contents IV 1. Introduction 1 2. Overview 2 3. Determinants of Economic Growth: A Cross-Country Analysis 5 3.1 The Basic Empirical Framework 5 3.2 Regression Results 10 3.3 Economic Growth of Latin America in Comparative Perspective 14 3.4 Growth Prospects for East Asia and Latin America 20 3.5 Extensions: Quality of Education and Income Distribution 22 4. Crisis, Adjustment and Sustained Growth 29 4.1. Defining Balance-of-Payment Crisis 29 4.2. Overview and Stylized Patterns of GDP Adjustment during Crisis 31 5. Determinants of the Output Cost of Crisis 42 6. Conclusions 48 References 50 Appendix Table Balance-of-Payment Crises in Latin America and East Asia, 1970-1999 54 Tables and Figures (in body of text) Table 1. Economic Growth in East Asia and Latin America 3 Table 2. Summary of Key Variables by Region, 1970-75 and 1995-2000 periods (Unweighted average) 9 Table 3. Cross-Country Panel Regressions for Per Capita GDP Growth Rate 11 Table 4. Contributions to Growth Differentials Between East Asia and Latin America, 1970-2000 (percent, annual average) 15 IV Table 5. Contributions to Growth Differentials of Latin American Countries Relative to the East Asian Average, 1970-00 (percent, annual average) 19 Table 6. Growth Prospects for East Asia and Latin America, 2001-2010 21 Table 7. International Tests in Mathematics and Science 24 Table 8. Effects of Gini Index on Growth and Other Variables 27 Table 9. Incidence of Balance-of-Payments Crises, Over the 1970-1999 Period 31 Table 10. Macroeconomic Indicators during Balance-of-Payments Crises in East Asia and Latin America 35 Table 11. Determinants of Real Output Cost from Balance-of-Payments Crises 46 Figure 1. Income Distribution, in the 1990s 26 Figure 2. GDP Growth Rates during Currency Crises 32 Figure 3. Adjustments of GDP Growth Rates in East Asian Crises 33 Figure 4. Adjustments of GDP Growth Rates during the 1997 in East Asian Crisis 34 Figure 5. Adjustment of GDP Growth Rates during Latin American Crises in the 1980s 40 Figure 6. Adjustment of GDP Growth Rates In Latin American Countries in the 1990s 41 V Growth and Adjustment in East Asia and Latin America† José De Gregorio Jong-Wha Lee 1. Introduction The determinants of economic growth, the policies that affect it and ways to spur it are all tremendously important issues. The well-being of the population and prospects for poverty reduction are intimately related to economic growth. The issue of potential growth is at the very heart of recent discussions on stabilization policy. Whether a central bank should attempt to cool down the economy or not, and the state of the actual fiscal impulse, will depend crucially on the view of the rate of growth that can be sustained without creating inflationary pressures. During the past four decades, growth rates in the world have varied greatly. The four East Asian tigers—Hong Kong, China, Republic of Korea (hereinafter, Korea), Singapore, and Taipei,China—grew extremely rapidly, at an average of over 6.0 percent a year in per capita terms between 1960 and 2000. On the other hand, many countries in Latin America recorded less than 1.0 percent growth during the same period. The high growth of East Asian countries, compared to the poor performance of Latin American economies leads directly to the question of what fundamental factors explain such differences, and what should be done to spur growth. In this paper we assess the evolution of growth in the two regions and attempt to explain the poor performance of Latin America relative to East Asia. Using cross-country growth regressions, we find that traditionally important growth factors such as investment, population growth, and the quality of human resources explain almost half of the difference in per capita GDP growth between East Asia and Latin America. Economic policy and institutional factors, such as rule of law, government consumption, macroeconomic stability and the degree of openness explain the other half. The occurrence of balance-of-payments crises also contributed to lower growth in Latin America relative to East Asia. We extend the discussion on growth determinants to the role of income distribution and the quality of education. Although those variables do not enter satisfactorily into growth regressions, due to collinearity or the lack of available data, we provide evidence suggesting that they may help to explain why the regions have different institutions and policies. † We appreciate the comments given by Raimundo Soto, John Weiss, and seminars participants at the Asian Development Bank Institute (ADBI)—Inter-American Development Bank (IDB) conference, the Central Bank of Chile and Univesitat Pompeu Fabra. Part of this work was prepared while Jong-Wha Lee was visiting the Central Bank of Chile. He gratefully acknowledges financial support from the Central Bank of Chile, and the ADBI. The views expressed are those of the authors and do not necessarily represent those of the institutions the authors are affiliated with. Correspondence: [email protected]; [email protected]. 1 Comparing the experiences following currency crises in East Asia and Latin America, we find that the adjustment process is in general consistent with the stylized vpattern observed in all crisis episodes. The mean growth rates tend to hit bottom at the time of the crisis or one year afterward, and then to return to the pre-crisis trend rate during the two or three years following the crisis. However, output losses have been very severe in some recent crises such as the East Asian crisis of 1997. Therefore, it is important for an economy to reduce its vulnerability to crises, and to avoid severe output losses once it is struck. By examining the patterns of adjustment from previous crisis episodes, we identify the factors that help countries avoid a large decline in growth during a crisis and recover to the pre-crisis potential growth path. It appears that adequate international liquidity, real exchange rate depreciation and a sound banking system play a critical role in avoiding severe shocks from a crisis. A good external environment also speeds up recovery, limiting the cost of crisis. We have also found that an expansionary monetary policy may be relevant in dampening crisis cost, but found no significant effects for fiscal policy. The paper follows in five sections. In Section 2 we present an overview of East Asian and Latin American growth over the past forty years. In Section 3, using crosscountry regressions, we explain what have been the critical factors behind Latin America’s low growth performance relative to East Asia, and also discuss prospects for the future. In Section 4, we analyze the patterns of adjustment from previous crisis episodes, and compare the experience of crisis episodes in East Asia and Latin America. In Section 5 we investigate the factors that help countries avoid severe output losses following crisis and to return more quickly to the pre-crisis potential growth path. Finally, Section 6 concludes. 2. Overview Compared to East Asia, Latin America’s growth performance has been disappointing. Table 1 presents the details of growth in the sample of 21 Latin American countries and 9 East Asian economies we analyze in this paper.1 All averages are constructed without weighting by size, thus giving the same weight to all countries.2 1 The East Asian nine economies are: People’s Republic of China, Hong Kong, China, Korea, Singapore, Taipei,China, Thailand, Malaysia, Indonesia, and the Philippines. The 21 Latin American countries are listed in table 1. 2 We use GDP data from Penn-World Tables version 6.1, as described in Summers and Heston (1991) and Heston, Summers, and Aten (2002). The selection of countries was made on the basis of data availability for regressions in Section 3. 2 Table 2. Summary of Key Variables by Region, 1970-75 and 1995-2000 periods (Unweighted average) All countries (N=85) East Asia (N=9) Latin America (N=21) 1970-1975 Period Per capita GDP growth 0.025 0.048 0.021 Per capita GDP in 1970 5472 2922 4273 Investment/GDP 0.190 0.216 0.145 Fertility rate in 1970 4.9 4.8 5.5 Schooling in 1970 1.21 1.37 0.83 Life expectancy in 1970 64.0 64.8 65.1 Government consumption /GDP 0.094 0.050 0.102 Rule-of-law index 0.560 0.611 0.381 Inflation 0.127 0.105 0.202 Democracy index 0.516 0.346 0.479 Openness -0.006 0.324 -0.147 Terms of trade -0.021 0.003 -0.009 Balance-of-payments crisis 0.26 0.22 0.29 1995-2000 Period Per capita GDP growth 0.018 0.025 0.011 Per capita GDP in 1995 9205 11291 5301 Investment/GDP 0.168 0.270 0.150 Fertility rate in 1995 3.2 2.2 3.2 Schooling in 1995 2.43 3.17 1.81 Life expectancy in 1995 68.9 71.8 70.7 Government consumption /GDP 0.077 0.071 0.106 Rule-of-law index 0.707 0.781 0.559 Inflation 0.080 0.044 0.114 Democracy index 0.680 0.500 0.732 Openness -0.022 0.689 -0.207 Terms of trade -0.012 0.006 -0.012 Balance-of-payments crisis 0.24 0.56 0.14 Notes: The sample consists of the 85 countries used in the regressions in Table 3. Per capita GDP levels and growth rates are based on 1996 international (purchasing power parity adjusted) prices, based on the World Tables 6.1, as described in Summers and Heston (1991) and Heston, Summers, and Aten (2002). Schooling data is the average years of male secondary and higher schooling from Barro and Lee (2001). The investment ratio is the ratio of real investment (private plus public) to real GDP, based on the World Tables 6.1, averaged over the period. The government consumption measure is the ratio of real government consumption (exclusive of spending on education and defense) to GDP, based on the World Tables 6.1. The rule-of-law index, expressed on a zero- to-one scale, with one being the most favorable, is based on the International Country Risk Guide’s maintenance of the rule of law index. The inflation rate is the growth rate over each period of the consumer price index. The democracy index, expressed on a zero-to-one scale, with one being the most favorable, is based on the indicator of political rights compiled by Freedom House. The openness variable is the ratio of exports plus imports to GDP, filtered for the estimated effects on this measure from the logs of population and area. The growth rate of the terms of trade is the change of export over import prices over the period. The balance-of- payments-crisis variable is described in the notes to Table 7. The nine East Asian economies are: PRC, Hong Kong, China, Korea, Singapore, Taipei,China, Thailand, Malaysia, Indonesia, and the Philippines. The 21 Latin American countries are listed in table 1. 9 Our regression of specification (1) applies to a panel set of cross-country data over six five-year periods from 1970 to 2000, corresponding to the periods 1970-75, 1975-80, 1980-85, 1985-90, 1990-95, and 1995-2000.5 The dependent variables are the annual growth rates of real GDP per capita over the same six five-year periods. Some previous studies use cross-section data in which each country has only one observation. The approach based on the panel data set seems to consider more information than what is available from time series variations within each country. One concern in the empirical specification is that any effect from contemporaneous explanatory variables may reflect a reverse causation from GDP growth to the explanatory variable. For example, the relationship between contemporaneous investment and growth may be because high growth causes high saving. This problem, however, can be solved by adopting an instrumental-variables estimation technique. We estimate this system of the six equations using three-stage least squares.6 The instrumental-variable technique controls for the possible simultaneity problem when Zi—the control variables—are endogenously determined. Instruments are mostly lagged values of the independent variables (see the notes to Table3). We use prior colonial status (Spanish or Portuguese colonies and other colonies) as instruments for inflation rate in the instrumental-variable technique as in Barro (1997). In order to control for the possible reverse causation from lower growth to higher frequencies of balance-of-payments crisis, we use the ratio of international reserve to monthly imports at the beginning of each five-year period as an instrument for balance-of-payments crisis. 3.2 Regression Results Table 3 shows the regression results using the basic framework of equation (1) and the explanatory variables just described. The three-stage least squares technique is applied to a data set for 85 countries. 5 We do not include the 1960s period in the regression because the currency-crisis variable is only available from 1970. 6 The estimation gives countries equal weights but allows for different error variances in each period and for a correlation of these errors over time. Some studies suggest estimating panel growth regressions by the fixed-effects estimation technique, considering for an unobservable country fixed effect. However, the fixed-effects technique eliminates information from cross-section variations (see Barro, 1997, pp.36-39). Temple (1999) discusses other statistical problems concerning the estimation and interpretation of crosscountry growth regressions. 10 Table 3. Cross-Country Panel Regressions for Per Capita GDP Growth Rate (1) (2) (3) (4) Log (per capita GDP) -0.0236 (0.0036) -0.0251 (0.0036) -0.0270 (0.0039) -0.0224 (0.0036) Investment/GDP 0.0723 (0.0272) 0.0560 (0.0274) 0.0558 (0.0270) 0.0497 (0.0280) Log (total fertility rate) -0.0180 (0.0058) -0.0151 (0.0060) -0.0153 (0.0064) -0.0132 (0.0060) Male upper-level schooling 0.0021 (0.0017) 0.0029 (0.0017) 0.0031 (0.0018) 0.0019 (0.0017) Log (life expectancy) 0.0546 (0.0209) 0.0653 (0.0214) 0.0614 (0.0237) 0.0661 (0.0225) Government consumption/GDP -0.0723 (0.0272) -0.0722 (0.0239) -0.1068 (0.0267) -0.0646 (0.0238) Rule-of-law index 0.0178 (0.0074) 0.0179 (0.0075) 0.0184 (0.0084) 0.0161 (0.0075) Inflation rate -0.0284 (0.0080) -0.0129 (0.0090) -0.0077 (0.0090) -0.0144 (0.0091) Democracy index 0.0556 (0.0183) 0.0599 (0.0188) 0.0562 (0.0212) 0.0555 (0.0190) Democracy index squared -0.0456 (0.0171) -0.0472 (0.0175) -0.0387 (0.0196) -0.0422 (0.0179) Openness measure 0.0072 (0.0045) 0.0086 (0.0046) 0.0112 (0.0049) 0.0038 (0.0046) Growth rate of terms of trade 0.0312 (0.0229) 0.0346 (0.0233) 0.0558 (0.0270) 0.0307 (0.0234) Balance-of-payments crisis (contemporaneous period) — -0.0165 (0.0053) -0.0168 (0.0058) -0.0161 (0.0051) Lagged balance-of- payments crisis — — 0.0061 (0.0056) — Group of 9 East Asian countries — — — 0.0106 (0.0056) Group of 21 Latin American countries — — — -0.0033 (0.0041) No. of countries 85 85 85 85 No. of observations 464 464 391 464 Notes: The system has six equations, corresponding to the periods 1970-75, 1975-80, 1980-85, 1985- 90, 1990-95, and 1995-2000. Column 3 omits the 1970-1975 period in order to include the lagged value of the balance-of-payments-crisis variable. The dependent variables are the growth rates of per capita GDP. Data on GDP are from Penn-World Tables version 6.1. The log of per capita GDP, the average years of male secondary and higher schooling, and the log of life expectancy at age one are measured at the beginning of each period. The ratios of government consumption and investment to GDP, the inflation rate, the total fertility rate, the growth rate of the terms of trade, and the democracy index are period averages. The rule-of-law index is the earliest value available (1982 or 1985) in the first three equations and the period average for the other equations. The openness variable is the period average. Estimation is by three-stage least squares. Instruments are the actual values of the variables for schooling, life expectancy, openness, and the terms of trade; dummy variables for Spanish or Portuguese colonies and other colonies (which have substantial explanatory power for inflation); lagged values of the log of per capita GDP, the government consumption ratio, and the investment ratio; and the initial values for each period of the rule-of-law index and democracy index. In the first two equations, the instrument for the rule-of-law indicator is its values for 1982 or 1985. The initial values of foreign reserve-import ratio are used as an instrument for balance-of-payments crisis. Individual constants (not shown) are included for each period. Standard errors of the coefficient estimates are shown in parentheses. 11 Column 1 of table shows the result of the basic regression without including the balance-of-payments crisis dummy variable. Column 2 includes as an independent variable the balance-of-payments crisis dummy. Although the two columns show a similar pattern of results, substantial differences arise for inflation and schooling variables. The estimated effect of inflation on growth becomes much smaller when we include the balance-of-payments crisis variable. This may reflect the strong positive correlation between inflation and balance-of-payments crisis. On the contrary, the schooling variable becomes more significant in column 2 where the balance-of- payments crisis variable is added. Since the balance-of-payments crisis variable itself enters very significantly, we focus on the results of column 2. The results provide strong evidence for conditional convergence: the coefficient on the log value of initial GDP in column 2 is highly significant, and the estimated coefficient is -0.025 (standard error = 0.004). Thus, a poor country with a lower initial income level grows faster, controlling for the variables that influence the steady-state level of income. Specifically, the coefficient implies that a country with half the income level of another will grow 1.73 percentage points (=2.5%*ln(2) ) faster. The investment rate and fertility variables have strong effects on the growth rate. The estimated coefficient on investment rate is positive and statistically significant at the 5% level. The coefficient 0.056 (s.e. = 0.027) implies that a one standard deviation increase, equivalent to 8.3 percentage point in the ratio of investment to GDP in the 1995-99 period, is associated with an increase in the growth rate of about 0.5 percentage points per year. The estimated coefficient on the logarithm of fertility rate is strongly negative, -0.015 (s.e. = 0.006), implying that an increase of 0.51 (the variable’s standard deviation) in the fertility rate in 1995 lowers the growth rate by about 0.8 percentage points per year. The result of column 2 shows that human resource variables have a significantly positive effect on economic growth. The educational attainment variable, which is measured by an average year of secondary and tertiary schooling of the male adult population, has a positive effect on the growth rate: the estimated coefficient, 0.0029 (s.e. = 0.0017), is statistically significant at the 10% critical level. The mean and standard deviations of the schooling variable are 2.4 and 1.5 per year respectively for 1995. Therefore, the coefficient indicates that one standard-deviation increase in the secondary and higher schooling raises the growth rate of per capita income by about 0.4 percentage points per year. The logarithm of life expectancy at age one—a measure of health attainment—is highly significant in the regression: the estimated coefficient 0.065 (s.e. = 0.021) implies that an increase in life expectancy by 0.13 (the standard deviation of the log of life expectancy) in 1990 is estimated to raise the growth rate by about 0.9 percent per year. We find clear evidence that institution and policy variables play a significant role in determining economic growth. The government consumption variable has a significantly negative impact on growth: an increase in government consumption ratio by one percentage point reduces growth by 0.07 percentage points a year. In the sample, a one standard deviation of 5.5 percentage points over the 1995-99 period decreases the growth rate of per capita income by about 0.4 percentage points per year. The rule of law index has a strong positive effect on growth, indicating that countries with more effective protection of property and contractual rights tend to have 12 higher growth rates. The estimated coefficient, 0.018 (s.e. = 0.008) implies that a one standard deviation increase of 0.22 in this index (on a scale of 1.0) in the 1995-99 period is associated with an increase in the growth rate of about 0.4 percentage points. The openness variable appears to be positively associated with the growth rate. The estimated coefficient 0.0086 (s.e. = 0.0046) is significant at the 10% level. An increase of an economy’s openness by 0.4 (its standard deviation) over the 1995-99 period is estimated to raise the growth rate by about 0.4 percentage points per year. The regression result confirms the non-linear relationship between democracy and growth, as found by Barro (1997). The coefficients of the indicator of democracy and its square terms are positive and negative respectively, and both are statistically significant. The pattern of coefficient values indicates that growth increases with political freedom when the level of democracy is low, but decreases with it once the society has attained a certain level of political freedom. The estimated coefficients in column 2 imply that the switch occurs at a level of democracy of 0.635. Both the Latin America and East Asian regions on average were below this critical value in the 1970s. However, in the 1995-99 period, Latin America’s average level of democracy, 0.732, slightly exceeded this critical level. Column 2 shows that the effect of inflation on economic growth is negative but statistically insignificant. The estimated coefficient, -0.013 (s.e. = 0.009), implies that an increase in the average rate of inflation by one standard deviation of 9.7 percent over the 1995-99 period lowers the growth rate by 0.1 % per year. Note that the coefficient is less than half the value of column 1, where inflation has a greater impact on growth. As we saw earlier, the problem is the correlation between balance-of-payment crises and inflation. The regression result shows a less significant effect of the terms of trade change on per capita GDP growth. The estimated coefficient on the growth rate of the terms of trade is 0.035 (s.e.= 0.023), indicating that countries with favorable terms of trade shocks by one standard deviation of 0.039 in the 1995-99 period grew by about 0.1 percentage point per year above other countries. The balance-of-payments crisis turns out to have a strong, negative effect on economic growth. The estimated coefficient of the balance-of-payments crisis variable is -0.017 (s.e.= 0.005), indicating that a balance-of-payments crisis shock lowers the growth rate by 1.7 percentage points per year. Column 3 of Table 3 adds the lagged effect of a balance-of-payments crisis. The result confirms those of Barro (2001). The retardation of growth by a balance-of- payments crisis does not persist into the subsequent five-year period. In fact, the effect of a balance-of-payments crisis on economic growth in the following five-year period turns out to be positive but statistically insignificant. Therefore, a balance-of-payments crisis reduces income permanently, but has no permanent effects on growth. Table 3 also shows the result of the regression with the inclusion of regional dummies. Column 4 shows that the Latin American dummy variable has a statistically insignificant coefficient, while the East Asian dummy is marginally significant at the 10 percent level. It is interesting to note that earlier empirical studies found a significant and negative “Latin American dummy” (Barro, 1991), which in the current empirical framework becomes insignificant, indicating that the explanatory variables included in the right-hand side explain most of the poor performance of Latin American economies. 13 However, the point estimates, although small in magnitude and statistically insignificant, still indicate that in addition to the variables included, Latin America has lower growth rates than average, and East Asia has higher growth rates than average. Even controlling for the two regional dummies, the regression shows that most of the explanatory variables are still significant and have estimated coefficients of the same magnitude, compared to those in column 2 of Table 3. 3.3 Economic Growth of Latin America in Comparative Perspective The results of the cross-country regression allow us to analyze the growth performance of the Latin American countries relative to that in other regions. We compare the growth performance of Latin America to the best performance of East Asia. Average per capita growth rates for the nine economies in the East Asia region were 5.6%, 5.1% and 4.3% over each decade in the 1970-2000 period, while those for the 21 Latin American countries were 2.1%, -0.8% and 1.6% respectively. We use the point estimates of the parameters in the regression (2) of table 3 to carry out a simple “accounting” that breaks down the fitted values of growth rates for each country into the contributions from each of the explanatory variables. Although the residual errors in individual country growth rates are substantial, it is worthwhile to examine the differences in the explanatory variables that generate differences in the fitted growth rates. We then explore the sources of the differences in the fitted growth rates between East Asia and Latin America. Table 4 presents the results. The basic regression can account for a substantial part of the growth differences between the two regions. For the 21 Latin American countries, the predicted growth rate is 3.1 percentage points lower on average than that for East Asia over the whole period from 1970 to 2000. The actual difference was 3.6 percentage points, and therefore we can explain the bulk of the differences. It is interesting to note, however, that the larger difference occurs during the “lost decade” of the eighties. The predicted difference can be broken down separately into the contributions of the 12 explanatory variables. 14 Table 4. Contributions to Growth Differentials Between East Asia and Latin America, 1970-2000 (percent, annual average) Contributions to the difference in per capita GDP growth of East Asia relative to Latin America 1970-80 1980-90 1990-2000 1970-2000 Difference in Actual Growth 3.26 5.33 2.29 3.62 Predicted Growth 3.40 3.97 1.87 3.08 (100%) Initial Income 0.91 -0.05 -1.28 -0.14 (-4.5%) Investment Rate 0.43 0.65 0.73 0.60 (19.6%) Fertility 0.31 0.56 0.63 0.50 (16.3%) Human Resources 0.14 0.28 0.45 0.29 ( 9.4%) Schooling 0.17 0.20 0.34 0.24 Life Expectancy -0.03 0.08 0.11 0.05 Institutions and Policy 1.46 1.95 1.29 1.57 (50.8%) Government Consumption 0.40 0.32 0.22 0.31 Rule of Law 0.41 0.32 0.37 0.37 Inflation Rate 0.16 0.55 0.22 0.31 Democracy 0.04 0.12 -0.27 -0.04 Openness 0.45 0.65 0.75 0.62 Terms of trade 0.01 0.04 0.02 0.02 (0.7%) Balance-of-payments Crisis 0.14 0.54 0.04 0.24 (7.8%) Note: Our sample includes the nine East Asian economies and 21 Latin American countries listed in Table 1. The predicted per capita growth rate is based on the estimation result of column (2) in table 3. The relatively higher income level of Latin America compared to that of Asia in 1970 led to lower growth there in the 1970-90 period because of the convergence effect. However, this convergence effect became rather favorable to Latin America after 1980, when the income of East Asia surpassed it. Hence, the net convergence effect becomes negligible over the three decades from 1970 to 2000. 15 In 1970, Latin America had a slightly higher life expectancy and thus better conditions for growth than East Asia. But in general, Latin America had relatively poorer human resources—in terms of lower educational attainment and lower life expectancy. The regional differences have widened over time. The net effect of human resources contributed to about slower growth by about 0.3 percentage points in Latin America relative to Asia over the whole period. The investment rate and fertility had strong effects on Latin America’s performance relative to Asia, by lowering the per capita growth rate by about 0.6 and 0.5 percentage points per year respectively over the past three decades. Without this difference, Latin America would have had a level of per-capita income 25% higher after the thirty years ending in 2000. The institution and policy variables turned out to have a significant effect on differences in growth rates. The differences in growth may be due to low (human and physical) capital accumulation, or low productivity growth. The growth effects of institutions and policies can occur by reducing productivity and the speed of catch-up to the technological frontier, and changing the incentives for (physical and human) capital accumulation. The combined effect of the differences in the five policy variables— government consumption, rule of law, inflation, democracy, and trade openness— accounted for 1.6 percentage points slower growth in Latin America relative to Asia over the period from 1970 to 2000. The institution and policy variables contributed most to the difference in growth rates, registering 2.0 percentage points in the 1980-90 period. That is, during the debt crisis, policies and institutions deteriorated significantly in Latin America. As we emphasize again below, although external conditions could have led to a deterioration of internal policies and institutions, the poor growth performance, even in a period with negative external environment such as the eighties, can be traced to bad policies and institutions. Among the institution and policy variables, trade openness was the most important variable. Latin America’s relatively inward-oriented trade strategy accounted for slower growth of about 0.6 percentage points per year. Not only does it have lower trade share (GDP share of exports plus imports), but most of the countries are smaller in size and population than Asian countries, which further reduces effective trade openness. The higher inflation in Latin America also reduced growth by 0.3 percentage points relative to Asia over the whole period from 1970 to 2000. The negative effect of high inflation was more significant in the 1980s, lowering growth by 0.7 percentage points in Latin America relative to Asia. During the 80s, the average inflation rate in Latin America was 48.5%, while for our sample of 9 East Asian countries it was 2.6%. As discussed above, this effect does not include the likely detrimental effects of inflation on investment; however the evidence shows that the effects of inflation on investment are much smaller than those of inflation on productivity growth (De Gregorio, 1996). Government consumption and rule of law also contributed to the lower growth rate of Latin America, by 0.3 and 0.4 percentage points per year respectively over the three decades. By contrast, democracy turned out to play an insignificant role on growth difference between the two regions. On average East Asia’s low political freedom, far 16 lower than the “critical turning point,” has been relatively unfavorable to economic growth. Because democracy has a nonlinear effect on per capita growth, very low or high values tend to be detrimental to growth. In this respect, the democracy level for PRC, Indonesia and Haiti were very low, while Costa Rica, and Trinidad and Tobago were on the high side of the distribution. But for most East Asian and Latin American countries, the democracy level has not made a significant difference to growth rates. Table 4 shows that the effect of the relatively unfavorable terms-of-trade shock was also small in Latin America. This result casts doubt on the view that the problem of Latin America was due to its patterns of specialization, which faced a particularly unfavorable external scenario. Advocates of the import substitution strategy of Latin America in the sixties argued that the countries ought to pursue internal industrialization since the products they exported faced declining terms of trade. However, the evidence from our regression shows that the latter argument is wrong and that it is precisely openness, as part of good policies and institutions, that boosts fast and lasting growth. In addition, the external environment could explain part of the poor performance during the debt crisis. The largest difference between predicted and actual growth occurs in the 80s. This difference of 1.4 percentage points, even after we control for policies, institutions, terms of trade and balance of payments crisis, cannot be explained by the growth regressions. Of course, as we discuss in the next section, the output losses from currency crises do not only depend on external factors, but also on some internal factors such as initial conditions and policy responses. On the other hand, the balance-of-payments crisis contributed to lower growth in Latin America by about 0.2 percentage points than in East Asia over the whole period. It had the biggest effect in the 1980-90 period, explaining the growth differential of 0.5 percentage points. But in the 1990s, when East Asian economies also suffered from balance-of-payments crises, its contribution to the growth differential became negligible. Thus, while initial income and external conditions explain only moderate differences in growth rates, the major differences are produced by investment, human resources and the institution and policy variables. The traditionally important growth factors such as investment, fertility, and quality of human resources contributed significantly to the difference in per capita GDP growth between East Asia and Latin America. Moreover, relatively poor economic policies, such as trade protection, high inflation, high government consumption, and lack of good institutions, have been very important factors contributing to the relatively slow growth of Latin American countries during the past three decades. Table 4 focuses on the relatively poor performance of Latin American countries compared to East Asia. But there were also tremendous variations in growth performance among Latin American countries. While the best performing—Dominican Republic and Chile—grew by 3.2 and 2.4 percent per year during the 1970 to 2000 period, the worst performers—Nicaragua and Venezuela—registered negative growth rates of -2.7 and -1.7 percent. In addition, growth rates fluctuated greatly within each country. For instance, average per capita growth for Chile was only around 1.2 percent over the period 1970-90 but increased dramatically to 4.8 percent over the period 1990- 2000. On the contrary, with the exception of Philippines, the 9 Asian countries all had strong growth throughout most of those three decades without significant variations. 17 For this reason, we investigate the extent to which variations in the growth performance of individual Latin American countries can be attributed to the factors that explain the international growth variations, and in particular, the extent to which the variations are due to differences in domestic institutions and policies. We therefore extend Table 2 of the entire period, breaking down Latin America into 21 individual countries. Based on the regression result of column 2 in Table 3, we can assess how much of the variations in the growth performance of individual Latin American countries relative to that in East Asia can be attributed to each explanatory variable. Table 5 shows the effect of various factors on the difference between predicted and actual growth to 10 selected Latin American economies over the whole period from 1970 to 2000. For instance, the predicted growth rate for Chile is 2.6 percentage points lower on average than the East Asian region, while the actual difference was 2.1. The relatively higher income level of Chile compared to East Asia led to a drop of 0.5 percentage points in Chile’s growth due to the convergence effect. Investment represented a 0.5 percentage points increase, while fertility and human resources made no significant contribution to the growth differential. Chile’s sound rule of law accounts for the 0.2 percentage point edge over East Asia but other institutions and economic policies were relatively unfavorable to growth. The combined effect of the differences in the other four policy and institutional variables—high government consumption, high inflation, low levels of democracy, and low trade openness—accounted for 1.5-percentage-point slower growth in Chile relative to the group of nine East Asian countries from 1970 to 2000. Most of these indicators improved during the nineties, contributing to growth above the East Asian average, and only surpassed by PRC and Taipei,China over the whole decade. 18 Among the 44 countries that participated in the IAEP and/or TIMMS projects, students in the East Asian economies—PRC, Hong Kong, China, Japan, Korea, Singapore and Taipei,China—showed the highest achievements in mathematics. For example, in the IAEP mathematics test, PRC ranked first with an average score of 80.2, followed by Korea and Singapore. In contrast, Brazil, the only participating Latin American country, came last, with an average score of 34.7, following Mozambique. Among the 39 countries that participated in the TIMSS, Singapore, Korea, Japan, and Hong Kong, China were the top four in mathematics, with average scores ranging from 56.4 for Hong Kong, China to 60.1 for Singapore. In contrast, Colombia, the only participating Latin American country, performed significantly more poorly with a mean of 36.9, placing it second to last after South Africa. The results are also favorable for Asia in science tests, for which Asian students performed much better than Latin Americans. Although evidence on the quality of schooling is still scarce, there is a very clear gap between Latin America and Asia, which adds to the deficiencies in the quality of human resources we already discussed in the previous sections. Another area where the differences between Latin America and Asia are evident, although not included in the independent variables in our regressions, is income distribution. Figure 1 shows Gini coefficients for Latin American and Asian countries, with Japan and the United States shown for comparison. The data are taken from the World Development Report of the World Bank, for the closest year, which in most cases is between 1996 and 1998. There are several problems that make it difficult to make cross-country comparisons. For example, differences depend on whether the unit of analysis is the household or individual, whether income is measured before or after tax, and whether the surveys refer to income or expenditure. However, given all of those caveats, the conclusion is undisputable: inequality in Latin America is much greater than in Asia, and as we argue below it can explain differences in human resources, policies and institutions between the two regions. 25 Figure 1. Income Distribution, in the 1990s 0 10 20 30 40 50 60 70 Japan Korea Indonesia PRC US Thailand Costa Rica Peru Philippines Panama Malaysia Venezuela Hong Kong, China Mexico Chile Colombia Brazil Gini Coefficient Source: World Bank, World Development Report. The relationship between income distribution and growth has recently become a hot topic. Theoretical discussions often predict negative effects of inequality on growth (Alesina and Rodrik 1994, and Persson and Tabellini 1994). Most cross-country empirical studies also find support for a negative relationship between income inequality and growth (Alesina and Rodrik 1994, and Perroti 1996). However, some recent studies based on panel-data estimation find a positive relationship (Li and Zou 1998, and Forbes 2000). The main problem in carrying out cross-country empirical investigations is the quality and comparability of the data measured, with small differences often resulting in large differences in the estimated relationship between inequality and growth. We have investigated the effects of inequality on growth using our panel framework. Our measure of income inequality is the Gini index. The data come from the UNU/WIDER—UNDP World Income Inequality Database (WIID), which extends the data set of Deininger and Squire (1996). The first row of Table 8 reports the estimated coefficient on the Gini index when it is added to the systems in row 2 of Table 3. The overall sample size for the panel regressions decreases from 464 to 277, because many fewer observations of Gini coefficients are available than for the full sample considered in Table 3. In the system, the Gini value for around 1970 appears in the equation for growth from 1970 to 1975, and so on. The five-year lagged values of the Gini coefficients are added to the list of instruments. 26 Table 8. Effects of Gini Index on Growth and Other Variables Dependent variable Gini index (1) Growth rate regression -0.001 (0.018) (2) Fertility (log) 1.335 (0.146) (3) Government consumption/GDP 0.143 (0.036) (4) Secondary school enrollment -0.800 (0.098) (5) Rule-of-law index -0.869 (0.126) Notes: In the growth rate regression in row (1), the Gini index is added to systems in column 2 of Table 3. The Gini value around 1970 appears in the equation for growth from 1970 to 1975, and so on. The five-year lagged value is added to the list of instruments. The regressions for the other dependent variables come from the systems of the six five-year periods from 1970 to 1999 for each variable. The log of per capita GDP and the Gini index are included as independent variables. Estimation is by three-stage least squares. Instruments are the lagged values of the log of per capita GDP and the Gini index. The estimation result shows that there is no significant impact of Gini coefficients on economic growth. The estimated coefficient, -0.001 (s.e.=0.018) is essentially zero. Thus, with the other explanatory variables considered in growth regressions held constant, differences in income distribution have no significant effect on subsequent economic growth. Although income inequality has no direct impact on growth, additional effects can arise from the influence of inequality on the explanatory variables. One of the effects suggested by previous studies involves the impact of income distribution on fertility. Row 2 of Table 8 shows the estimation result for a panel system in which the log of the fertility rate is the dependent variable. In this system, the explanatory variables include the log of per capita GDP and the Gini index. The lagged values of the log of per capita GDP and the Gini index are used as instruments. The result confirms a strong positive impact of inequality on fertility. In theories based on political economy arguments, inequality affects government expenditures and thereby growth. In unequal societies, there are more incentives for redistributive politics (Meltzer and Richard, 1981). Row 3 of Table 8 shows direct consideration of a panel system in which government consumption ratio is the dependent variable. We find a significant influence from the Gini index. Another channel by which income inequality influences growth is educational attainment. Poor families with borrowing constraints are not able to invest in their children even when the returns on education are very high. They have problems sending their children to school even under free schooling, since they often need income from their children’s employment. This occurs relatively less in more equal societies, given the same level of income, since the parents are able to pay the costs of education. A more equal distribution enables more households to send their children to school. Row 4 of Table 8 confirms that income inequality has a strong negative impact on secondary school enrollment. Lower secondary school enrollment leads to smaller secondary 27 educational stock in time, and consequently has an adverse impact on economic growth. Thus, income distribution affects growth through the human capital channel. We also find a strong negative impact of income inequality on institutional quality. Row 5 of Table 8 shows the estimation result for a panel system in which the log of the rule-of-law index is the dependent variable. We find a significantly negative impact of the Gini index on the rule of law. Political economy considerations can also help to explain why corruption, rule of law, and institutional quality in general are weaker in more unequal societies. Overall, we find substantial evidence that inequality affects growth indirectly by influencing fertility, government consumption, education and rule-of-law. Consider as an example the estimated coefficient of 0.143 on the Gini coefficient in row 3 of Table 8. This point estimate implies that an increase in the Gini coefficient by 0.1 (its standard deviation), that is 10 percentage points, raises the government-consumption ratio by 1.4 percentage points of GDP. If we multiply this value by the estimated coefficient of the government-consumption ratio in the growth regression (-0.07 in column 2 of Table 3), we get -0.001. Thus, this indirect channel lowers economic growth by about 0.1 percentage point. Similarly, the point estimate of -0.87 of the Gini coefficient in the regressions for rule of law (row 5 of Table 8) and the estimated coefficient of the rule-of-law index in the growth regression (0.018 in column 2 of Table 3) imply that an increase in the Gini coefficient by 0.1 leads indirectly to a decrease in growth rate by about 0.16 percentage point through deteriorating institutional quality. This evidence suggests that although income distribution has no significant direct effect on economic growth in our regressions (row 1 in Table 8), inequality may be detrimental to economic growth by increasing distortions, weakening institutions and reducing the quality of human resources. More research needs to be done to establish the definite connections, since up to this point we have seen some very suggestive correlations. In addition, the accounting exercises overall show that although the gap in the growth rate between Latin America and East Asia will narrow over the next decade, it will still remain substantial due to the differences in investment, fertility, schooling, as well as certain policy variables such as government consumption, rule of law, and inflation. Therefore, Latin America must do more to improve the investment rate, fertility, schooling, and institutions. Two important policies to achieve this objective involve increasing public saving and expanding educational enrollment. An increase in public saving will contribute to raising the investment rate, containing pressures on government consumption, tax distortions, and high inflation rates. An increase in educational enrollment, particularly at the secondary level, will help to lower fertility and increase educational attainments. However, improving public finance and education investment is not easy. As we analyzed in Section 3.2, high government expenditures and low educational enrollment are to a certain extent the outcome of unequal income distribution. Latin American countries have a more unequal income distribution than East Asian countries, and the evidence on the determinants of income distribution show that this gap cannot be closed in a short period of time. For example, improvements in education take time 28 to pass through to a large share of the labor force (see, e.g., De Gregorio and Lee, 2002). 4. Crisis, Adjustment and Sustained Growth In the previous section, we find that a balance-of-payments crisis reduces growth rates in the years close to the crisis. Such crises have contributed to a difference in growth performance between East Asia and Latin America of 0.25 percentage points per year over the 1970-2000 period. This is not minor, since it represents about 40% of the difference explained by investment rates, and is nearly equivalent to the difference explained by human resources. Moreover, if East Asia had avoided the severity of the1997 crisis, a much larger part of the regional growth differences would have been attributed to balance-of-payments crises. Hence, it is important for economies to reduce their vulnerability to crisis and thereby avoid the deleterious effect of financial turmoil on economic growth. Once an economy is hit by a crisis, the necessary adjustments must be made quickly so that it can return to sustained growth. There has been a considerable literature, starting with Frankel and Rose (1996), on identifying the determinants of currency crises, or at least searching for good indicators that can help to predict the occurrence of crises. Our purpose is different, and we take a different route as we are interested in determining the output costs after a balance-of-payments crisis has occurred and which factors can help to alleviate these costs. In the following section, we analyze the patterns of adjustment observed in previous crisis episodes, and compare the experience of adjusting from currency crises in East Asia and Latin America. We investigate the factors that help countries avoid severe output losses following crises and return more quickly to the pre-crisis potential growth path. 4.1. Defining Balance-of-Payment Crisis To examine the nature of adjustments from crises, we first need to define what a crisis actually is. A balance-of-payments crisis is typically defined as an event when an index of exchange market pressure exceeds a certain threshold. Several alternative indicators and methods have been used in the literature to identify the dates of currency crises. Frankel and Rose (1996), Milesi-Ferreti and Razin (1998), Barro (2001) and Park and Lee (2002) use the nominal depreciation rate as the index and date the emergence of each crisis as the time when the index increased sharply over an exogenous threshold rate of depreciation common to all countries. Severe speculative pressure does not lead to large depreciations if the authorities successfully defend the currency by intervening in the foreign exchange market. Hence, Eichengreen, Rose and Wyplosz (1995), Kaminsky and Reinhart (1999), and Glick and Hutchison (2001) use an alternative indicator of currency pressure by combining depreciation rates with additional variables such as foreign reserve losses and domestic interest rate. A balance-of-payments crisis is considered to have occurred if the composite indicator increased above a threshold level in terms of the countryspecific moments. For example, Kaminsky and Reinhart (1999) construct an indicator of currency pressure by a weighted average of monthly nominal depreciation rate and 29 monthly percentage change in foreign reserves, with weights such that the two components of the indicator have an equal size in terms of sample volatilities. A balance-of-payments crisis is then judged to have occurred in any year when the change in the indicator of currency pressure for any month of that year exceeded three standard deviations above the mean of the indicator over the sample period for each country.8 This procedure is also subject to a potential problem, since it implies that, if the form of the distribution was equally normal but the mean and standard deviations varied across countries, the expected number of crises would be the same for all countries. Our indicator of currency crises combines two approaches. We define a balance-of-payments crisis as an episode identified either by the former or the latter approach. For the former approach, we judge a country to have a balance-of-payments crisis if it experienced a nominal currency depreciation of at least 25 percent in any quarter of a specific year and the depreciation rate exceeded that in the previous quarter by a margin of at least 10 percent. In applying the criterion in the latter approach, we count those episodes in which the indicator of currency pressure for any month of the year exceeded three standard deviations above the mean of the indicator, provided that either the monthly nominal depreciation rate or percentage change of reserve loss exceeded 10 percent. We apply a window of three years to isolate independent crises. That is, a balance-of-payments crisis occurring up to three years following an initial crisis is counted as a continuation of the same crisis rather than a new episode of crisis. Applying this procedure, we identify 260 independent currency crises for 130 countries over the period from 1970 to 1999. Of them, 221 occurred in developing countries. Latin American countries suffered more balance-of-payments crises than East Asia—55 compared to 14—in the 1970-1999 period. Table 9 is a summary for the patterns over time and across regions of our definition of balance of payment crises. 8 Kaminsky and Reinhart (1999) try to exclude the orderly large-scale nominal depreciation in hyperinflation countries by separating the sample observations in which inflation in the previous six months was higher than 150 percent. 30 Table 9. Incidence of Balance-of-Payments Crises, Over the 1970-1999 Period Period All countries All developing countries East Asia 9 Latin America 21 1970-1999 260 221 14 55 1970-75 26 15 2 6 1975-79 29 21 1 7 1980-84 46 39 4 12 1985-89 47 46 1 14 1990-94 72 62 1 13 1995-99 40 38 5 3 Notes: A balance-of-payments crisis is defined by combining two criteria. A crisis is judged to occur in the year when a country experienced a nominal currency depreciation of at least 25 percent in any quarter of the specific year and the depreciation rate exceeded that in the previous quarter by a margin of at least 10 percent. A crisis is also identified to have taken place in the month of a year when an indicator of currency pressure—a weighted average of monthly nominal exchange depreciation and monthly foreign reserve loss—exceeds three standard deviations above the mean of the indicator over the sample period for each country, provided that either the monthly nominal depreciation rate or percentage change of reserve loss exceeds 10 percent. A crisis that does not take place at least three years after the latest crisis is counted as a continuation of the initial crisis rather than an independent crisis. 4.2. Overview and Stylized Patterns of GDP Adjustment during Crisis Figure 2 shows the movements of real GDP growth rates before and after currency crises during the1970 to1999 period. We show the movements of the sample mean of the GDP growth rate at the onset of the balance-of-payments crisis; in each of the preceding six years; and in each of the ensuing eight years. For comparison, we include a straight line in the figure indicating the average GDP growth rate during the ‘non-crisis’ period that was not subject to a crisis in that specific year. We use the sample of developing countries. 31 Figure 2. GDP Growth Rates during Currency Crises -0.02 0.00 0.02 0.04 0.06 0.08 0.10 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 +7 +8 Years Before/After Crisis No crisis All Crises East Asia 9 L. America 21 The growth rates clearly demonstrate a v-type pattern during the period bordering the occurrence of a crisis. The growth rates over the six to two preceding years are comparable to those of non-crisis periods, at about 3.9 percent in the sample of developing countries.9 But thereafter, the growth rate starts to decline sharply and reaches the trough at the crisis year. This v pattern of real GDP adjustment over the period before and following the crisis is broadly consistent with the findings in Gupta et al. (2002), and Park and Lee (2002). At the time of the trough, the growth rate of all crisis-hit developing countries is about 1.1 percent. However, the GDP growth rate recovers to its pre-crisis trend level two or three years after the outbreak of crisis. The quick recovery of GDP growth rates to their trend level in crisis-hit countries is consistent with the fact that balance-of-payments crises slow growth rates only temporarily, as we found in the previous section. Thus, when an economy is hit by a crisis, it tends to recover its potential trend growth rate quickly. Therefore, the level of the trend growth rates seems to play a critical role for the adjustment pattern. However, recessions caused by balance-of-payments crises must incur permanent output losses because the post-crisis growth rates do not exceed the pre-crisis period averages, and for this reason we cannot think of them as output movements around a long-term trend, but rather as periods with permanent costs in terms of output and welfare. Figure 2 shows that the adjustment process in both East Asia and Latin America, as can be inferred from the movements of the growth rates, is broadly consistent with the stylized v-pattern observed from the all crisis episodes. The mean growth rates hit the bottom at the time of the crisis or one year afterward, and then quickly recov over the following two years. A similar v-type pattern of adjustment in both East Asia and Latin American crisis episodes suggests that despite the conventional sense of 9 In order to have data updated up to 2002 (and forecasts for 2003), we use GDP growth rates from IMF, World Economic Outlook database as of April 2002, rather than per capita GDP from Summers-Heston. The non-crisis average of per capita GDP growth rates is 2.1% over the sample period. 32 structural differences between East Asia and Latin America, the pattern of adjustment from crisis is similar. Although the v-shape pattern is similar, the adjustment is much sharper in East Asian than in Latin American crises. The deeper initial contraction following crises in the East Asian countries must be attributed to the severity of the 1997 crisis. Figure 3 exhibits the adjustment patterns from crises in East Asian economies. It confirms that the contraction of real income in countries that suffered from crisis in the 1990s was much larger than it was in the previous decades. The five countries that were most affected by crisis—Thailand, Indonesia, the Philippines, Malaysia, and Korea—suffered a sharp decline of real income. In 1998, the GDP growth rates of the five crisis-hit Asian countries plunged from the pre-crisis average of 7.0% to negative figures ranging from –13.1% in Indonesia to –0.8% in the Philippines (see Figure 4). Figure 3. Adjustments of GDP Growth Rates in East Asian Crises -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 -6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6 +7 +8 Years Before/After Crisis 1970s 1980s 1990s 33 Figure 4. Adjustments of GDP Growth Rates during the 1997 in East Asian Crisis -0.14 -0.10 -0.06 -0.02 0.02 0.06 0.10 0.14 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Year INDONESIA KOREA MALAYSIA PHILIPPINES THAILAND It is interesting to note the similarities across Asian countries during the crisis. As shown by Figure 4, all followed the same qualitative pattern, and even the figures are broadly alike. This indicates that this was a regional phenomenon. Although it is out of the scope of this paper to measure fundamental vs. self-fulfilling crises, as well as the impact of contagion, the similarity suggests that in addition to similar fundamental domestic problems, as discussed below, there were common external shocks, such as the deterioration of business confidence or erratic behavior of financial markets, that explain the high correlation in the evolution of GDP. Several factors contributed to the deeper contraction in the 1997 East Asian crisis. One important one is that the crisis was in some aspect a severe liquidity crisis caused by investor panic (see Radelet, Sachs and Lee 2001). After Thailand fell victim to a currency crisis in July 1997, there was rampant speculation that other East Asian countries might be unable to avoid currency devaluation and debt default. When foreign investors panicked, capital flows into the countries quickly reversed. Over the pre-crisis period from 1993 to 1996, the five countries had capital-account surpluses of about 7% of GDP on average. This surplus fell to 0 in 1997 and into a deficit of 4.8 percent of GDP over the following four years (see Table 10). Thailand experienced a sudden reversal of about 19 percent of GDP between 1996 and 1997. This reversal was even more pronounced in terms of private capital inflows, reaching an adjustment that amounted to about 23 percent of GDP (from a net inflow of 11 percent in 1996 to a net outflow of 12 percent in 1997). Official creditors offset only a small fraction of the reversal of private capital. In Korea, the capital account went from a surplus of 4.6 percent of GDP to a deficit of 1.9 percent. Net bank lending outflows amounted to 8.3 percent of GDP. 34 addition, the links between the banking system and the corporate sector increased financial fragility.13 Figure 6. Adjustment of GDP Growth Rates In Latin American Countries in the 1990s -0.18 -0.15 -0.12 -0.09 -0.06 -0.03 0.00 0.03 0.06 0.09 0.12 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003f Year ARGENTINA BRAZIL CHILE MEXICO COLOMBIA The main difference between Chile and East Asia was the evolution of inflation, which was at an annual 20% or more, except in the year of recession when it was 13 percent, illustrating that the decline in the rate of inflation in the year of the crisis was linked to the fixed exchange rate and the decline in output rather than to a reduction in long-term inflation. It is interesting to note that Chile in the 1982 crisis and Mexico in the 1994 crisis demonstrated a v-shape pattern of adjustment most similar to that in Asia, and similar to the broad evolution of GDP shown in Figure 1. One could argue that the private origin of the massive crisis, with a strong initial fiscal position, may have allowed for an orderly, although costly, resolution of the financial crisis. With the exception of the case of Colombia, short term debt was often a problem in Latin America, where short term liabilities were from 2 to 6 times larger than reserves, and hence led to serious liquidity problems. In addition, the sharp depreciation that theses countries experienced after the crisis may have helped them to recover more quickly, as was seen in Asia. Indeed, comparing the debt crisis in Latin America, Edwards (1995) shows that between 1982 13 It is also interesting to note, as emphasized by Edwards (1999), that Chile had capital controls in 1982, more stringent than the famous controls of the nineties, but this did not allow it to avoid the currency crisis. 41 and 1987, Chile’s depreciation was about 90%, while in Argentina, Brazil and Mexico it was on the order of 20-30%. In addition, in the late 1980s the inflationary problems of Argentina and Brazil led them to adopt heterodox disinflation policies that were essentially price controls without fiscal adjustment, and as expected, they were a complete failure, a final touch to the lost decade. Latin America during the nineties has suffered from currency crises, but there have been different timings, characteristics, and causes. Therefore, when examining the growth experience, it is difficult to attribute them to a common cause, such as the implementation of reforms, as might be argued by those who oppose reform, or to am unfavorable external environment. The first crisis of the 1990s started with Mexico’s tequila crisis, which then dragged Argentina into a financial crisis. Argentina was able to maintain convertibility, at a severe cost in terms of output. Both countries followed a v-shape adjustment. However, Argentina recovered by 1998, but concurrently with the Asian crisis experienced another deterioration of economic conditions, which led to the collapse of convertibility in 2001. Brazil and Chile were also hit hard by external conditions. Brazil, after struggling to avoid a currency crisis, experienced one after the Russian crisis— which is singled out as the most proximate cause of Brazil’s crisis—at the end of 1999. If we consider currency depreciation and loss of reserves, Chile did not experience a currency crisis in the late 1990s. However, the interest rate defense of strong currency pressures in 1998 would, in a broader definition, have been close to a currency crisis. In Argentina, as a result of the appreciation of the dollar in the international markets and the depreciation of Brazil’s real, there was growing pressure on Argentina’s dollarpegged peso, which finally collapsed in December 2001. Since the late 1990s, all Latin American countries have been struggling to return to higher growth, but as shown in the previous section, their growth prospects are definitely lower than Asia’s. By contrast, the quick turnaround of the East Asian economy after the 1997 crisis has brightened the region’s economic prospects. As the recovery continues, these economies have seemed to return to sustained growth. Despite the impressive record of the recovery, however, not everyone is confident about East Asia’s future prospects. Long-term sustained growth is no guarantee that the countries will not be susceptible to a severe crisis in the future. Asian countries still face challenges in increasing their resilience to crisis, especially in the financial sector. 5. Determinants of the Output Cost of Crisis In this subsection, we formally investigate the factors that determine the magnitude of output losses accompanying a crisis. We attempt to determine which kind of initial conditions, policy reactions and external conditions help to reduce the cost of crisis. 42 We measure the output cost of a crisis by the cumulative loss in output growth during the period from the year when the crisis begins until the time when output growth returns to its trend.14 As a first step, we define the period of crisis as three years—the crisis year and the following two years. Therefore, the output cost of each crisis episode is calculated as follows: )(cos 2 0 it i growthGDPtrendtoutput + = −= ∑, where t indicates the year in which a crisis occurs. The variable trend represents the trend GDP growth rate. This trend is calculated using the average GDP growth rate during the “non-crisis” period, that is a year in which the country was not subject to a crisis and the preceding two years. For each country, this is the average rate of growth during the non-crisis year, where crisis consists of a 3-year period.15 According to this definition, the average output cost per crisis from 1970 to 1999 was about 5.8 percentage points for an average developing country. Note that this figure is consistent with our estimations in Section 3, where we found that a crisis reduces growth by 1.7% per annum during a five-year period. Since we identify the crisis for three years, this is about 5% lower output. However, the standard deviation of the output cost of crisis is considerable: 11 percentage points. This implies that the crisis episodes had different characteristics in terms of origin, evolution, and policy responses. In fact, not every crisis episode was associated with output loss: it turns out that output was expansionary in about 30 percent of all crisis episodes. A well-known case is the crisis of the European Monetary System in 1992, where most countries hit by the crisis grew strongly afterwards. During the same period from 1970 to 1999, the average output cost was about 8.9 percentage points on average for Latin American crises and about 10.4 percentage points for an average East Asian country. 14 This measure should be viewed as an indicator of output losses associated with a crisis, but not necessarily caused by the crisis. Lower output growth during recession may trigger a balance-of- payments crisis, which in turn aggravates the recession. IMF (1998) and Eichengreen and Bordo (2001) have used the same output cost to measure the cost of crises. In the latter case, the output loss is measured until output returns to trend, whereas we fix a three-year period. The figures, however, yield similar results for the output cost of a crisis, since the recovery time is about two to four years. Alternatively, the output cost of a crisis can be measured by the cumulative loss in output during the crisis years. In order to construct this, we need to have the measure of potential output. If we assume that the potential output of each country increases from the pre-crisis equilibrium by the rate of the non-crisis trend growth, we can construct the potential output over the three years following a crisis and measure the output cost of a crisis by the sum of the log differences between the potential output and potential output over the three years. The estimation results based on this cost measure, which can be provided upon request by the authors, are very similar to those reported in this section. 15 For an alternative measure of the trend growth rate, we have considered the predicted growth rate coming from specification 1 of Table 3 in Section 3. The estimation results based on this measure are broadly similar to those reported in this section. However, the sample size is much smaller. 43 We believe a large number of factors led to the differences in output cost in the crisis-hit countries. The nature of the shock, initial conditions, external environments, and policy responses must influence the behavior of output following a crisis. For the purpose of our empirical investigation we classify these factors into two broad categories— pre-crisis and post-crisis factors. As pre-crisis factors we include: (i) pre-crisis GDP growth rate, (ii) international liquidity, as measured by an indicator of reserve adequacy, and (iii) banking sector soundness. The real GDP growth rate prior to the crisis can provide information about the degree of imbalance in the economy; in particular we focus our attention on GDP growth compared to its trend rate of growth. If an economy has a high GDP growth rate with respect to trend prior to a crisis, it is likely to be in an over-expansion, and therefore the crisis will not only have traditional disruptive effects on economic activities, but will also help to bring output closer to normal levels. For example, previous studies have shown that lending booms and excessive credit expansion during a pre-crisis period tend to deepen the post-crisis recession and reduce growth (De Gregorio and Guidotti, 1995, and Sachs, Tornell, and Velasco, 1996). In addition, high-growth countries may be more vulnerable to severe shocks, thereby experiencing a larger output loss following a crisis. Adequate international liquidity is also important in dampening the shock of a currency crisis. A lack of foreign reserves has often resulted in a sharp liquidity crisis and a deep contraction of real output. As a measure of international liquidity, we use the size of foreign reserves relative to the broad money supply (M2). We also consider the ratio of foreign reserves to short-term foreign debt as an alternative measure. A sound banking sector is also important in preventing illiquidity. Vulnerability in the banking sector often magnifies a shock. In particular, when a currency crisis is associated with a systematic banking crisis, it becomes more costly. This refers to the “twin crises” that exacerbate the costs of currency crises. Indeed, one mechanism through which a currency crisis harms the economy is the balance sheet effects on the corporate and banking sectors due to maturity and currency mismatches. These were very common in the East Asian crisis of 1997 and in many of the Latin American experiences. Hence, we include a banking crisis variable as an important factor affecting the severity of the balance-of-payments crisis. The data on banking crises are compiled from Caprio and Klingebiel (1996), Demirguc and Detragiache (1998), and Glick and Hutchison (2001), who document the episodes of bank insolvencies based on both quantitative and qualitative criteria. The dummy for banking crises is an episode where a currency crisis in a specific year is preceded by a banking crisis during the previous two years or followed by a banking crisis in the subsequent two years. Looking at post-crisis factors, there are a number of characteristics that can affect the costs of a crisis. Following the discussion of the East Asian and Latin American experiences in the previous section, we consider the important post-crisis 44 factors to include: (i) world GDP growth, (ii) real exchange rate depreciation, and (iii) macroeconomic policies. The global economic environment is important for the post-crisis adjustment of crisis-hit countries. Strong world growth has a positive effect on export growth and helps crisis-hit countries to recover quickly, through improvements in terms of trade and increasing market access for their exports. The size of the initial real exchange depreciation following a crisis can also influence export and output growth in the postcrisis period. It is important to distinguish nominal from real depreciation, and indeed, it must not be assumed that letting the currency weaken will necessarily result in a real depreciation. However, evidence suggests that following a currency crisis, the passthrough from exchange rate to inflation is relatively small, but some suitable domestic conditions are necessary.16 Macroeconomic adjustment policies implemented by the government as part of crisis management are also important for the post-crisis recovery of real output. Fiscal and monetary policies can play a critical role in fostering recovery, and of course, they have been at the center of many policy discussions on the appropriate mix for making efficient adjustments to currency crises. For example, during the Asian crisis, a major issue of discussion was the role of expansionary fiscal policy and the need to tighten monetary policy at the beginning of the crisis to establish credibility and avoid excessive currency depreciation.17 It is out of the scope of this paper to analyze the very short-term policy reaction to a crisis, since this framework only allows us to shed light on the overall policy stance during the period and its impact on recovery. To the extent that the relevant data are available, we carry out an empirical assessment of the factors determining the output costs of balance-of-payments crises. Using the complete data set from 1975 to 1998, we set up a basic equation as follows (the subscript s~v, means average for the period from s to v): (2) output ttttt ttttt ebalancebudgetgrowthmoney ondepreciatirealgrowthpartnerstradecrisisbanking +⋅+⋅+ ⋅+⋅+⋅+ ++ ++− 2~72~6 52~42~23 ' ββ βββ ttttt Mreserveforeigngrowthcrisispret ⋅+−⋅+= −−−+ 124~2102~ ]2/[cos βββ where e is a random disturbance term. Table 11 shows the estimation results. We have also included dummies for the decades of the 1980s and 1990s in order to control for unobserved period-specific shocks. 16 Borenzstein and De Gregorio (1999) for example show that the pass-through is smaller in countries that had low inflation previous to the currency crisis. 17 See Fischer (1998) and Stiglitz (2002) for contrasting views. 45 Table 11. Determinants of Real Output Cost from Balance-of-Payments Crises (1) (2) (3) Pre-crisis GDP growth (average t-2~t-5) 0.660 (0.253) 0.804 (0.259) 0.594 (0.261) Foreign reserve/M2 (t-1) -0.146 (0.042) -0.154 (0.043) Foreign reserve/short-term debt (t-1) -0.0025 (0.0014) Banking crisis (t-2~t+2) 0.076 (0.020) 0.066 (0.023) 0.081 (0.021) Trade partners’ GDP growth (t~t+2) -2.735 (0.963) -2.571 (0.963) -2.916 (0.991) Real exchange rate depreciation ( t ) -0.035 (0.016) -0.028 (0.017) -0.036 (0.016) Real money supply growth (t~t+2) -0.129 (0.066) -0.157 (0.071) Growth of real money supply-to-GDP ratio (t~t+2) -0.046 (0.071) Budget/GDP (t~t+2) 0.311 (0.297) 0.133 (0.325) 0.214 (0.301) Dummy for 1980s 0.011 (0.027) 0.017 (0.034) 0.016 (0.027) Dummy for 1990s -0.036 (0.031) -0.041 (0.039) -0.031 (0.032) No. of crisis episodes 81 73 81 Notes: Robust standard errors are in parentheses. The dependent variable is a measure of output cost from a crisis, calculated by summing the differences between the trend GDP growth rate and GDP growth rates over the three years including the crisis year and the following two years. Column 1 shows that all explanatory variables enter with the expected signs. We find a strong and statistically significant positive relationship between the pre-crisis GDP growth and the output cost of the crisis. This may imply that a country with high growth relative to its trend—an economy that is overheating prior to a crisis—tends to have a larger decline in GDP growth over the three years following the crisis. The estimated coefficient (0.660, s.e.=0.253) implies that a one-percentage-point higher growth rate in the pre-crisis period increases the accumulated output cost by about 0.7 percentage points in the crisis-hit economy. As expected, the crisis tends to eliminate, although not completely, this “excess growth.” Adequate international liquidity prior to a crisis, which is measured as a ratio of foreign reserve to money supply, is estimated to decrease the output cost of the crisis. The estimated coefficient is negative and statistically significant at the 5 percent level. The estimated coefficient (-0.146, s.e.=0.042) implies that a 25 percentage point 46 increase (standard-deviation) in the foreign reserve-to-M2 ratio lowers the output cost by about 3.6 percentage points. A dummy variable for the occurrence of a banking crisis is positive and statistically significant. The estimated coefficient (0.076, s.e. = 0.020) implies that when a balance-of-payments crisis is accompanied by a banking crisis, the output loss increases substantially, by 7.6 percentage points. In rough terms, we can say that on average a twin crisis is almost like two currency crises in terms of the output costs. The results also show that several factors in the post-crisis period have strong effects on output cost. The world growth variable—an average of the GDP growth rates of a crisis-hit country’s trading partners weighted by its trade share—turns out to have a significant effect on crisis cost. The estimated coefficient (-2.74, s.e.=0.96) implies that a one percentage point increase in world GDP growth rate per year is associated with an approximate 2.7-percentage-point decline in output cost of the crisis over the three subsequent years. Note that the size of the coefficient in terms of annual growth rate is closer to one, since a one percentage point rise in world growth reduces the costs in terms of output by 2.7 percentage points over three years. The results also confirm that a real exchange depreciation helps to reduce the output cost. 18 The estimated coefficient of the real exchange depreciation variable is negative and statistically significant at the 5 percent level. The estimated coefficient (-0.035, s.e.=0.016) shows that a real exchange depreciation of 70 percentage points (its standard deviation) lessens the decline of GDP growth rate by about 2.4 percentage points over the three subsequent years. Column 1 includes average growth rates of the real money supply as the macroeconomic policy variable. This variable turns out to have a significant effect on the output cost of crisis. The estimated coefficient (-0.129, s.e.=0.066) implies that a 10 percentage point increase in real money supply over the post-crisis period leads to a 1.3- percentage-point drop in the output cost. Hence, post-crisis expansionary monetary policy helps to alleviate the output cost of a crisis. In contrast to the positive and significant contribution of monetary policy, fiscal policy turns out to have a negligible effect on the cost of a currency crisis. In column 1, we add the budget balancing variable as a measure of fiscal policy stance. The estimated coefficient for the budget balancing variable is positive, suggesting that a fiscal deficit tends to lower the output cost, but in a statistically insignificant way. The dummy variables, although significant, conform to the presumption that the crises of the 1980s were more costly. Comparing the parameter estimates, the higher cost of a crisis in the 1980s is about one half of a percent of GDP. Column 2 uses the ratio of foreign reserves to short-term debt as an alternative measure of international liquidity. This measure also has a negative relationship to the output cost of the crisis, and the estimated coefficient is marginally statistically significant at the 10% level. Column 3 concerns the possible endogeneity problem of monetary growth. The negative correlation between real money supply growth and output cost may come from the positive effect of output growth on money demand, and an accommodation from the supply-side, rather than the reverse. In order to avoid this problem, we use the growth of 18 Note that the variable is the rate of real depreciation in the first year of the crisis, which is less subject to the problem of reverse causality. 47 the real money-to-GDP ratio (M/PY), the inverse of velocity, as a measure of expansionary monetary policy. We still find that expansionary monetary policy tends to lower the output cost of crisis. However, the estimated coefficient (-0.046, s.e.=0.071) is not statistically significant.19 In sum, the results suggest that adequate international liquidity, real exchange rate depreciation and a sound banking system play a critical role in preventing severe shocks from a crisis. A good external environment also speeds up recovery, limiting the cost of the crisis. We have also found some relevant effects from expansionary monetary policy in dampening crisis cost, but no significant effect of fiscal policy. However, this does not provide straightforward implications for macroeconomic policies. In many cases, countries attempting to conduct expansionary macroeconomic policies may not be able to do so. For example, the procyclicality of fiscal policy in Latin America occurs due to the inability of governments to borrow during bad times (Perotti and Gavin, 1997). Along the same lines, a central bank that lacks credibility may end up being unable to adopt an expansionary, non-inflationary monetary policy. Therefore, the result may also be due to the institutional capability of countries to implement countercyclical macroeconomic policies. What our results show is that in past experiences of currency crisis, monetary policy has been more effective in alleviating output costs. 6. Conclusions We have compared the experiences of growth performance and macroeconomic adjustment in East Asia and Latin America. We have focused our analysis on nine East Asian economies and 21 in Latin America. The cross-country regression highlights the role of investment, human resources, fertility, and institutional and policy factors, which have made the potential growth rate of East Asia higher than that of Latin America over the 1970-2000 period. We find that the convergence effects due to differences in initial income or external shocks to terms of trade did not play a quantitatively important role in explaining lower growth in the Latin American region. This finding is important since the rationale for the inward looking trade strategy followed by Latin America since the 1960s was based on the belief that it would suffer from deteriorating terms of trade, an assumption which proved to be wrong. In this respect, among policy and institutional factors, the most important difference between the two regions has been the larger degree of openness of in East Asia. Balance-of-payments crisis shocks have contributed to differences in the growth performance across regions, although these differences shrank during the 1990s as a result of the Asian crisis of 1997. Low investment rates in Latin America were also important in explaining the differences with East Asia. In addition, high fertility rates, the low quality of human resources, mainly in terms of schooling, high government consumption, and low indices 19 The fiscal balance can be also procyclical and thereby subject to a possible endogeneity problem. But the positive estimated coefficient of fiscal balance implies that this endogeneity problem is of little concern. We also use the ratio of real government consumption to GDP as a measure of fiscal policy stance. Since government expenditure is considered to be less elastic to the cycle than revenues, the potential endogeneity problem is smaller. The government consumption variable also turns out to be statistically insignificant. 48 for the rule of law, are all factors that have contributed to the low growth in Latin America vis-à-vis East Asia. The quality of education, measured by results on international test scores, is also substantially lower in Latin America. However, we cannot estimate its quantitative impact due to a lack of long time series, and this is an issue that deserves further examination. Another important difference across regions is income distribution. We report some preliminary evidence which suggests that some of the differences in institutions and policies are the results of differences in income distribution. This is the case for fertility, government consumption, secondary school enrollment and rule of law. Using the regression results, we estimate growth prospects for both regions and find that the differences should decrease, due to a large extent to an unfavorable convergence effect for Asia, but also due to progress made in institutions and policies. There still remains a significant gap in terms of investment, fertility, schooling and openness. In any event, Latin America is predicted register average per-capita growth of 2.3% during this decade, while East Asia is expected to grow at 3.8%. We also analyze the pattern of recession and recovery from balance-of- payment crises and find that the adjustment process in both regions is broadly consistent with the stylized v-pattern we observe in all crisis episodes around the world. However, one characteristic differentiating some Latin American countries from Asia is the tendency of the former to escape from one crisis only to fall into another. This recurrence of balance-of-payment crises still requires further explanation. One exception has been the case of Chile, which in the 1980s suffered a crisis very similar to those of Asian economies in 1997. Although the origins may have been different, the crisis occurred simultaneously with a severe banking crisis. However, these two crises occurred under a sound fiscal policy, and the exchange rate declined sharply. In addition, the Chilean economy, like the Asian economies, was sufficiently open to recover through export expansion. We have looked at a sample of 81 episodes to identify factors that could reduce the output cost of crisis. As an external factor we find that a favorable international environment reduces the costs. At the other extreme we find that sound banking is very important, since the output costs of a twin-crisis (balance of payments plus banking) is about twice that of a balance-of-payment crisis alone. In addition, maintaining international liquidity before the crisis, as measured by international reserves as a ratio of M2, reduces the cost of crisis. In terms of policy response, our results indicate that a real exchange rate depreciation and expansionary monetary policy help in the recovery, while fiscal policy has no effects. Further work should endeavor to uncover the conditions that generate effective expansionary macroeconomic policies to increase resilience to an unfavorable external environment and to improve the quality of the adjustment. For example, in terms of achieving real exchange rate and relative price adjustments, real wages must generally fall. The wage agreements achieved in Korea were important in achieving this goal. Again, we suspect the differences in income inequality across regions may affect the likelihood of reaching such agreements. This could also help to explain the lack of full adjustment in Latin America, and pave the way for muddling through the next crisis. 49 References Alesina, A. and D. Rodrik (1994). “Distributive Politics and Economic Growth,” Quarterly Journal of Economics, 109(2): 465-490. Barro, R. (1991). “Economic Growth in a Cross Section of Countries,” Quarterly Journal of Economics, 106(2): 221-247. Barro, R. (1997). Determinants of Economic Growth: A Cross-country Empirical Study, Cambridge MA, MIT Press. Barro, R. (1999). “Human Capital and Growth in Cross-Country Regressions?” Swedish Economic Policy Review,6(2). Barro, R. (2001). “Economic Growth in East Asia Before and After the Financial Crisis,” NBER Working Paper 8330, June. Barro, R. and J.W. Lee (1994). “Sources of Economic Growth,” Carnegie-Rochester Conference Series on Public Policy, 40: 1-46. Barro, R. and J.W. Lee (2001). “International Data on Educational Attainment: Updates and Implications,” Oxford Economic Papers 53(3). Barro, R.J. and X. Sala-Martin (2004). Economic Growth, 2nd edition, Cambridge MA, MIT Press, forthcoming. Borensztein, E. and J. De Gregorio (1999). “Inflation and Devaluation after Currency Crises,” mimeo, IMF. Borensztein, E., J. De Gregorio, and J. W. Lee (1998). “How Does Foreign Direct Investment Affect Economic Growth?” Journal of International Economics, 45. Borensztein, E., and J.W. Lee (2002). “Financial Crisis and Credit Crunch: Evidence form the Korean Firm Level Data,” Journal of Monetary Economics 49(4), 2002. Burnside, C., M. Eichenbaum and S. Rebelo (2001). “Prospective Deficits and the Asian Currency Crises,” Journal of Political Economy, 109(6): 1155-1197. Caprio, G. and Klingebiel, D. (1996). “Bank Insolvencies: Cross-Country Experience,” Policy Research Working Paper No. 1620, Washington DC: World Bank. Corbo, V. and P. Rojas (1993). “Investment, Macroeconomic Instability and Growth: The Latin American Experience,” Revista de Análisis Económico, 8(1): 19-35. De Gregorio, J. (1992). “Economic Growth in Latin America,” Journal of Development Economics, 39: 59-84. De Gregorio, J. (1996). “Inflation, Growth and Central Banks: Theory and Evidence,” World Bank, Policy Research Working Paper No.1575. 50