University of Southern California Dissertations and Theses

Political economy of economic reform, quality of reform and economic performance: Reform matters, but consistent incentives matter more

(USC Thesis Other)

Political economy of economic reform, quality of reform and economic performance: Reform matters, but consistent incentives matter more

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content POLITICAL ECONOMY OF ECONOMIC REFORM, QUALITY OF REFORM AND ECONOMIC PERFORMANCE - REFORM MATTERS, BUT CONSISTENT INCENTIVES MATTER M O R E - by Junkyu Lee A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (POLITICAL ECONOMY AND PUBLIC POLICY) August 2004 Copyright 2004 Junkyu Lee R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UMI Number: 3145226 Copyright 2004 by Lee,Junkyu All rights reserved. INFORMATION TO USERS The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. ® UMI UMI Microform 3145226 Copyright 2004 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Dedication I dedicated this Ph.D thesis to m y beloved w ife Jungah, an d m y precious G od's gifts tw o d au g h ters C hristine, and Irene. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Acknowledgments I acknow ledge th a t w ith o u t the energizing guidance of D r. Jeffrey N ugent, Dr. K atada, an d D r.Chen, this thesis could n o t be achieved. A nd I especially ow e w o nderful w isdom an d insights to D r. Jeffrey N u g en t w ho is m y m entor. I appreciate m y fellow G raduate stu dents w ho have been helping m e navigate this journey of new challenge. I appreciate U niversity of Southern C alifornia Econom ics D epartm ent and Political Econom y and Public Policy P rogram to give m e an o p p o rtu n ity to finish this Ph.D. degree. I appreciate sincere prayers of m y parents and m y friends especially D aniel Yongjin Lee. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table of Contents A bstract xiii I. Reform m atters? A. Introduction 1 B. Building Theoretical Model: Moral hazard, Sequential Rationality, mean- variance utility function, expected income for the government 1. Moral Hazard 6 2. Dynamic Games: Credibility, Sequential Rationality, backward Induction 7 3. Introduction of the Political Economy Model 8 II. Political Economy of Export Support Financial Policy (ESFP) A. Definitions of Policy Loan 14 1. What is policy loan? 14 2. Changes of Policy Loans 15 B. Definitions, Forms, and Procedures: Export Support Financial Policy (ESFP) 1. Definitions, Formations and Procedures 21 (a) Foundations of ESFP (1) Contents of ESFP 2. Sources of ESFP and Support Procedures 23 3. The Forms of Loan Procedures 24 (a) L/C base loan process (b) Actual export volume base process 4. ESFFs Policy Procedures 25 5. The forms of ESFP 26 (a) Interest Rate (b) Loans of amount 6. Background of introducing the Export Support Financial Policy 27 C. Changes of ESFP with other factors: Pre- and Post-Structural Adjustment Periods 28 1. Pre-Structural Adjustment (1962-1971): Export Orientation Policy emphasized on the Light Industry 29 (a) International Political Economy: No pressure under GATT (b) A Relationships of Government and Firm: Strong Stackelberg Leader R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. V (c) A Relationships of the Government and Financial System: Monetary policy, BOK, Interest Rate Realization Act (d) The Analysis of the Export Support Financial Policies: 1962-71 (1) ESFP in the First and Second 5 year Economic Development Plans: 1962-71 (2) Characteristics of ESFP in the First Period (1962-1971) 2. Pre-Structural Adjustment (1972-1979): Export Priority Policy Period 42 (a) International Political Economy: Low Pressure (b) A Relationships of Government and Firm: Strong Stackelberg Leader (c) Government and Financial System (d) Export Support Financial System: 1972-1979 (1) The Analysis of the ESFP changes in 1972-79 (2) The Characteristics of ESFP Loans in the second period (1972-1979): 3. Structural Adjustment Period (1980-1987): Adjustment of Export Policy and Reform-initiation 55 (a) International Political Economy: high pressure (b) A Relationships of Government and Firms: (c) A Relationships of Government and Financial System (d) Export Support Financial System: lowering the export incentive parameter (1) The analysis of the ESFP changes: 1980-87 (2) The Characteristics of the ESFP Loans in the third period: 1980-87 4. Structural Adjustment & Openness and Globalization 1988-1997: Price reform Period 72 (a) International Political Econom y: high pressures to liberalize (b) A Relationships f Government and Firms (c) A Relationships of Government and Financial System (d) Changes in Export Support Financial System (1) The analysis of the ESFP changes: 1988-97 (2) The Characteristics of the ESFP Loans in the third period: 1988-97 D. Comparing of all periods 87 1. Government, International pressure and Government-Firm relationship 2. Changes in ESFP in different four periods R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. III. Empirical Analyses A. Conceptual Framework 89 B. Data 91 C. Investigating the inter-relationships in Structural Change and Simultaneity 1. Structural break in the incentive parameter and Analysis of Variance 95 (a) Structural Change in Incentive Parameter 94 (b) Firm's Commitment to exports in response to the incentive parameter change 99 2. Contemporaneous Relationships and Simultaneity 101 (a) Contemporaneous Relationships : only two variables ESFP Loans and Exports 101 99 (b) Two Single Equations and Simultaneity 102 (1) Two Single Equations: Relationships between ESFP Loans and Exports 102 (2) Two Single Equations: Relationships between Exports and ESFP Loans 104 (3) Polynomial Distributed Lag Model in the Dynamic Economic Model 105 (4) Simultaneity Issues and Hausman Specification Test 107 D. Investigating the inter-relationships in a Vector Autoregression (VAR) Model 1. Issues and Theory of Standard VAR(P) MODEL 110 (a) Alternative Specifications 114 (b) Ordering of Variables 115 (c) Trend Removal 117 (d) Level of Temporal Aggregation 116 2. A Estimation of a Standard VAR(P) 119 3. A standard VAR, Granger Causality Test with foreign pressures and export price index 125 4. Structural Vector Autoregression: Forecast Error Variance Decomposition 129 (a) SVAR: Identification and Choleski Decomposition 129 (b) Comparing Forecast Error Variance Decompositions 133 (1) FEVD based on lower triangular Cholesky decomposition 133 (2) Structural FEVD 139 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. (i) Effect of an innovation to Export Support Loan on Exports: Compare SFEVD of Before the Structural Adjustment and After the Structural Adjustment 139 (ii) Effect of an innovation to Exports on Export Support loan: Compare SFEVD of Before the Structural Adjustment and After the Structural Adjustment 140 IV. The Conclusion 141 References 153 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. List of Tables Table 2-1 Objectives of Policy Loans 16 Table 2-2 The trend of two policy loans to total policy loans (unit: %) 16 Table 2-2-1 The trend of the amount of policy loan in the Deposit Money Banks, unit: 100 mill won. 17 Table 2-3 Members of Financial Monetary Operation Committee in 1997 25 Table 2-4 General Economic Data 36 Table 2-5 Changes of Export Support Financial Policy 38 Table 2-6 Status of ESF Loans (100 millions Won, %) in the Deposit Money Banks 39 Table 2-7 Won per dollar for Export Financing, unit: Won 40 Table 2-8 The spread of interest between ESF and General Loans, % 40 Table 2-9 The features of ESFP in 1962-1971 41 Table 2-10 Sources of financing Unit: % 49 Table 2-11 General Economic Data 50 Table 2-12 Changes of Export Support Financial Policy 52 Table 2-13 Status of ESF Loans (100 millions Won, %) 52 Table 2-14 Won per dollar for Export Financing Loans Unit: Won 53 Table 2-15 The spread of interest rates between ESF and General Loans 53 Table 2-16 The Features of ESFP Loans in 1972-1979 55 Table 2-17 New Multipolarized Word Export Status, unit: percentage 57 Table 2-18 Share of the Policy Loans in the total loan 62 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 2-19 Economic Data 67 Table 2-20 Changes of Export Support Financial Policy 68 Table 2-21 Status of ESF Loans (100 millions Won, %) 69 Table 2-22 Won per dollar for Export Financing Loans 69 Table 2-23 The spread of interest rates between ESF and General Loans, % 70 Table 2-24 The Features of ESFP Loans in 1980-1987 71 Table 2-25 Firms' Net Purchase of the land and Total Loans 78 Table 2-26 Economic Data 81 Table 2-27 Changes of Export Support Financial Policy 82 Table 2-28 Status of ESF Loans (100 millions Won, %) 83 Table 2-29 Won per dollar for Export Financing Loans 84 Table 2-30 The spread of interest rates between ESF and General Loans, % 85 Table 2-32 The Features of ESFP Loans in 1988-97 86 Table 2-33 Pressures from International Institutions, and relationships with firms 87 Table 2-34 Comparisons of Export Support Loan Systems 87 Table 2-35 Comparisons of Interest rates and amount of loans 88 Table 2-36 The Analysis of the International Pressures, government leadership and policy tools 88 Data table History of the data 92 Table 3-1 Structural change in the incentive parameter 97 Table 3-2 Structural change in the incentive parameter in the GLO 97 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 3-3 Scheffe's test for the growth rates of the exports: ANOVA 99 Table 3-4 Growth rate of Exports in response to new ESFP loan structure 100 Table 3-5 The first differenced Loans and Exports 101 Table 3-6 The first differenced Exports and Loans 102 Table 3-7 ESFP loans, Exports, ESFP Nominal Interest Rates, Interaction Dummy 103 Table 3-8 ESFP loans, Exports, Spread of the two rates, Interaction Dummy 103 Table 3-9 The growth rates of Exports, Loans, Export Price Index, Wage Index, Interaction Dummy 104 Table 3-10 Polynomial Distributed Lag model: Lag = 4, polynomial = 1 105 Table 3-11 Polynomial Distributed Lag model: Lag = 4, polynomial = 1 106 Table 3-12 Hausman specification test of GLOHAT and LVHAT 109 Table 3-12-1 VAR(p) order selection criteria 122 Table 3-13 VAR(4) estimation 123 Table 3-14 VAR(4) with a foreign pressure exogenous variable 127 Table 3-15 Granger causality of export support loans and export when there is a foreign pressure 127 Table 3-16 Vector Autoregression: Loan, Export, Export Price Index 128 Table 3-17 First sample period: 1967.1-1984.4 134 Table 3-18 Second sample period : 1985.1-1996.4 135 Table 3-19 Two exogenous variables of EPI, DWI: First period sample: 1967.1- 1984.4 136 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 3-20 Two exogenous variables of EPI, DWI: Second period sample: 1985.1- 1996.4 138 Table 3-21 Structural FEVD before the break: Orderings: Loan, Exports, SOR 139 Table 3-22 Structural FEVD after the break: Orderings: Loan, Exports, SOR 140 Table 3-23 Structural FEVD before the break, Orderings: Export, Loan, SOR 140 Table 3-24 Structural FEVD after the break, Orderings: Export, Loan, SOR 141 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Box 2-1 Box 2-2 Figure 2-1 List of Figures The Sources of ESFP and Supply Channel 23 Export procedures and ESP supply procedures 24 Influencing Relationships between the government, banks and the firms 76 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Abstract This paper states a question of financial economic reform on firms used to moral hazard situation. In a frame of Political Economy which argues a heterogeneity of interests, we assume firms have been in moral hazard and government sets the incentive structure where the firms are operating. The idea of Stackelberg leader was introduced to deal with the sequential decision mechanism to model firm and government decision-makings. We argue a firm has utility maximization objective function and responds to an incentive structure - so called 'rew ard system' - which is instituted by the government. A government is a leader in policy-making and a firm is a follower in response to the policy decision of government. A firm is assumed to be risk-averse and a government is assumed to be risk-neutral. The specific target loan which is 'export support policy loan' was deemed as a reward to an export firm. This paper investigates how the political economy factors such as international pressure, and agenda changes influence a development policy making and how a representative firm responds to these changes. We expected a firm decreased its export commitment when a government eliminated the favorable incentive system. A firm might look for other opportunities to maximize objective function. A government who received international pressure of more open economy also is expected to respond to exogenous changes which will be on export loan policy-makings and this new institution is expected to influence firm's R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. x iv commitment to exports. Thus, the objective of this paper is to investigate how a firm and government make their decisions when they consider each other's decision as sequentially given. Theoretical moral hazard model was constructed. Empirically, we introduced various steps of econometric methods to prove inter relationships. First, we used the time series data of South Korea which has to be checked about the stationarity. Second, we tested whether there are structural changes in each single equation. As expected, dummy variable for interaction is significantly related to each other's decision on both before-the-break and after-the- break. Third, we consider an inter-relationship in a simultaneous equation system to be made sense. We found there was an endogeneity by Hausman Specification Test. Fourth, an endogeneity paves a way for us to consider a Vector Autoregression Model (VAR). Forecast Error Variance Decomposition (FEVD) explains how much a variance of exports is influenced by export support loans and export loan is influenced by export commitment on both before-the-break and after- the-break with other variables considered. As argued, econometric methods prove, after the structural break, that a firm and government decrease their commitments in this political economy model which assumes a heterogeneity of interests and they may cause non-optimal economic outcomes in a moral hazard construction in a course of economic reforms. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 I. Reform matters? A. Introduction Economic reform policy is different from traditional macroeconomic policy which pursues price stability and stable employment. The Reform pursues rational allocation of resources and enhancement of economic efficiency is expected to accompany. According to the various definitions of economic reforms, most of Developing countries have been implementing economic reforms encompassing a broad range of market-oriented policies during the 1980s and 1990s. Many Latin American countries experienced the economic crisis two times - one in 1980s and the other 1990s. East Asian countries experienced the currency crisis in 1997 and the recovery is on the way back since 1997. Especially South Korea is the one who has strongly adopted the economic reforms over a last few decades and experienced the crisis as other developing countries did. There are a lot of theories and researches which are trying to explain why the East Asian developing countries had to go through the crisis. My puzzle starts from what the problem of the focus only on price reform is. Because those developing countries followed the guides of neo-classical economic reform policy so far, if the guides are appropriate and the price reform of deregulation is a nicely equipped ideology, the question left is why these developing countries had to see the economic crisis. So what would be an R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 institution to achieve the sustainable growth? I want to argue this question is related to the one "what is the superior institution1 " which is non-market but actually market-supporting institutions2 . Even if the countries followed the market reform well, when the reform is lack of growth-promoting quality, the economy may face the crisis. I assume that one of the growth-promoting qualities of reform is a better incentive structure which gives a consistent motivation for economic decision makers to commit their full resources to growth. In this paper, I will introduce the political economy3 model of economic development in the stage of reform policy. Therefore, we will prove how importantly the different incentive systems given by government under international pressures play roles in determining economic commitments by firms which have been used to moral hazard situation. Depending on different incentive systems on different periods over the time horizons of economic development policies, it will be investigated that whether firms commits different effort to different incentive systems. 1 . According to Lin and Nugent (1995) institutions are a set of humanly devised behaviroal rules that govern and shape the interactions of human beings, in part by helping them to form expectations of what other people will do. This literature had the cross-national work that quantifies the growth- promoting effects of superior institutions. . Rodrik (1999) said market supporting institutions were the property rights, regulatory institutions. Institutions for macroeconomic stabilizations, institutions for social insurance, and institutions of conflict management. 3 . Drazen(2000), “ Were there no heterogeneity of preferences over outcomes, there would be no need for a mechanism to aggregate individual preferences into a collective choice. Similarly, were there no conflicts of interests whatsoever, the choice of economic policy would be that of the social planner maximizing the utility of the representative individual. It is heterogeneity of interests that is the basis of the political economy”, plO R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 The political economy model of reform here is on the bases of Moral Hazard, and Dynamic Games. Based on them, the theoretical model is introduced to capture how this heterogeneity of interests will determine the choices of the agents in response to the predetermined institutions. In the following chapter II, we will present how the political economy factors surrounding export support policy loan have been changed over the time periods. How international pressures appear in the firm-govemment relationships will be presented. At chapter III, we will present how the firm responded to the ESFP loan policy changes with various econometric methodologies. We will investigate whether there are structural changes in the incentive parameter and how it affects the exports commitments. We tested the unit root / nonstationarity for the data and made the data stationary to be used in the time series analysis. We used the method of traditional structural macro model - dynamic simultaneous equation system, polynomial distributed lags. We present the issues of simultaneity to deal with the vector autoregression. Finally, we will employ a VAR and Structural VAR model for the Forecast Error Variance Decomposition to find out how much the change in the variation of each variable which is explained by the variations of other variables with pre- and post- structural adjustment. Therefore, we will prove different incentive changes will R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 surely influence the firm's different commitment to exports which is the engine of economic growth in developing countries. We will see moral hazard firms commit less efforts to exports to unfavorable incentive system given by the government. The contributions of this paper are the followings: (1) in terms of methodology, to deal with the incentive changes of economic agents, time series methodology such as VAR with forecast error variance decomposition (FEVD) is introduced. Comparisons of FEVD before and after the structural break are a new way in determining the decisions of economic agents. There is a new method to explain changes of incentives of firms and government which are shown by time series econometric methods. And the structural VAR model is newly introduced to explain the inter-relationships of firm and government based on the economic theory. (2) in terms of economic theory; Two stage Game theory is put into the base of this model. It introduces moral hazard, sequential rationality, and political economy model to explain different decisions of firms and governments over the periods. First of all, in order to explain the different commitment of firms and government, it uses the different heterogeneous objective functions4 which are new approaches in the theory of economic reform policy. Firm's objective function used to be a profit 4 . By the terms of Laffont {The Thoery o f Incentives, 2002), this is called Political Economy approach. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 maximization type, but here I adopt the utility maximization function which may fit in the developing countries' model. (3) it allows the interactions between the domestic economic policy and international foreign pressure in the model and empirical analysis. It is shown that domestic economic policy and the decisions of economic decision-makers change responding to international pressure. This model is under the small open economy in the developing countries. B. Building Theoretical Model: Moral hazard, Sequential Rationality, mean- variance utility function, expected income for the government In order to capture the idea of economic reform policies and performance and quality changes of reform on growth under the changes of parameters, I devised5 and introduced the two theoretical models which involved moral hazard and dynamic games. It is the essential subject that how the firms react to the optimal incentive structure which is given by the government. I considered the hidden action of firms' commitments and the sequential rationality with dynamic games as the government is a Stackelberg leader. This moral hazard and Sequential Rationality over the time horizon would explain the firms' optimal commitments and how the government chooses the 5 . Following the mechanism design, hidden action problem and optimal decision mechanism, I introduced a theoretical model. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 optimal incentive scheme. This change of optimal incentive parameter involves the change of quality of reform regardless of big bang or gradualism of financial reform. Next two sections will show the backbones of moral hazard model and ideas of the dynamic games introduced here to explain the political economy between the private agents "Firms" and the government. 1. Moral Hazard Agency relationships arise whenever the person who undertakes an action (the firms) is not the same as the person who bears the consequences of that action (the principal: the state or the government). We assume the agent is the enterprises and the principal is the government. A representative firm has the utility function which is based on net income stream. Usually when we use the objective function of the firms, it is the net profit function. Differently from the commonly used moral hazard model, I used the utility function as a objective function of the firms. There are enough reasons to adopt the utility function as an objective function. In most of the developing countries, the owner of the company exercises the whole decision-making power. For example, the owner of the Korean Chaebols expanded the scope of the products based on the owner's own decision. When Samsung chairman decided to enter the car industry which the Samsung never did before in 1997, the chairman said "this car industry is my father's wishes. Therefore I do". Many cases show that the owners of the firms identify himself with the firms especially in the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 developing countries. Based on the behavior mechanism of most of the developing countries' enterprises, it would be better to have the Utility function instead of net profit function. 2. Dynamic Games: Credibility, Sequential Rationality, backward Induction Government and firms are doing this dynamic game two times. Both are assumed to be rational and complete information is assumed in a sense that the players' payoff functions are common knowledge. Perfect information is assumed in a sense that when the government and the firm have the historical data thus far once they do the game. So they remember what they chose in the previous games. This dynamic game6 makes the shortcomings of the moral hazard overcome with these assumptions. Credibility7 is the central issue in this dynamic game. They know each other they are rational not in just one stage but also over the time span. So firm will not believe the non-credible threats from the government and government will not believe the firms' non-credible decisions. They are sequentially rational. Multiple Nash Equilibrium may be possible, but the only Subgame Perfect Nash Equilibrium is the equilibrium associated with the backwards-induction outcome. 6 . K r e p s ,(1990), A course in Microeconomic Theory, 355-577 7 . Mas-colell, Whinston, and Green (1995), Microeconomic Theory, Dynamic Games, 267-301 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 8 When the government chooses its choice variable, it anticipates the reaction function of the firms and when the firm chooses its choice variable, it anticipates the reaction function of the government and includes the government's reaction function as exogenous decision in order to maximize its utility. With this dynamic games and moral hazard model, we would explain how the firms' decision to the performance depends on the optimal incentive scheme. 3. Introduction of the Political Economy Model In the overall model setting of the political economy in die economic development, we assume heterogeneity of interests is necessary for there to be political constraints. Heterogeneity plays out quite differently when addressed through the market. There are two economic decision-makers - one is representative firm and another is the government. Government is assumed to be principal and the firm is assumed to be the agent. Government sets up the incentive structure and confines the institutions where the firm maximizes the objective function. The government at the initial stage announced the economic development plan and firm sequentially observed the government's reward system in terms of specific target loans - export support loans. Thus the firm considers the government decision as given and follows her leadership. Government observes the firm's commitment to the economic agenda - especially in terms of exports - and determines the reward sequentially. The firm R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 9 considers the government reward commitment is sequentially credible. The government does, too. They do this dynamic game two times under the moral hazard8 model9 . We first introduced the firm's objective function and see how the firm determines the choice variables. Then we introduced the government's objective function and see how the government determines her choice variables. Firms' reward structure1 0 is, G = a + p Y (1) G: domestic credit1 1 allocated by the government Y: Firm's performance, for example the amount of export or the O utput a: fixed reward P: Strength of the incentive. This p tells how the change of quality of reform affects the commitments of the firms and the magnitude of P measures the strength of the incentives. Assume that export Y is a continuous function in input plus a white noise. We introduced the mean-variance function to explain how the risk and international pressure affect the firms' performance. Firm wants to maximize its expected utility function which includes the risk factor "r" and international pressure "a2 ". 8 . In the development stages, the government has been helping the industry in various ways and the firms have been used to this environment. 9 . Kreps, A Course In Microeconomic Theory, 577-614. Mas-colell, Whinston, and Green (1995), Microeconomic Theory, 477-510. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 10 The firms are assumed to be risk-averse. Max V = a + p F(G,L) - WL2 - rp2 cr2 (2) Where F is the production function, W is the wage rate, r is a risk factor, a2 is the variance of international pressure, and L is the Firm's choice variable to the higher commitment or lower commitment for the exports. First Order Condition with respect to L which is the firms' choice variable turns out, 5F/9L = P F l - 2 WL = 0 (3) From this (3) result, we can get the Second Order Condition with respect to P, We get, SL*/3p = -F l / p F ll - 2W > 0 (4) Therefore, when the government increases the incentive structure, the firm will do a high effort to increase the Exports or GDP. The key is that when the firm decides how much of input to increase the Exports, the firm accepts the Government's Optimal Incentive Scheme parameter ft as given. Firm includes the government's decision into its objective function as parameters. Next, we investigate the government objective function. Government indirectly influences firms' input commitments by manipulating the strength of incentives. 10. Equation (1) is given by the government initially. The firm observed this reward structure and makes a decision for the next stage. 11. Assume: export support loan R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 11 We assume that the government is risk-neutral. Government will maximize its objective function and is benevolent one who will take care of national income. W hat government does is to choose what is the optimal incentive structure - that is, choosing the level of p which tells the change of the quality of reform. Optimal contract will specify "P" that maximize government share of the expected output. Government makes its decision under three constraints - participation constraint, incentive constraint and BOP debt constraint. The objective function is, Max -a + (1-p) F(G,L) Subject to Participation Constraint: a+ PF(G.L)- WL2 - rP2cr2= UO1 2 Incentive Compatibility Constraint: L = L * (p, W, G) BOP Debt Constraint: k_bar = N(G(P), L*(P, W)) where kjbar > = 0. (5) From the First Order Condition, Government chooses p. When we assume the implicit function rule applies well, We have, P = P* (w,r, a 2 ) (6) When we differentiate the (6) with respect to risk, international pressure and debt - with the Bordered Hessian Matrix1 3 showing, we get the result of second order condition for constrained maximization, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 12 dp/dr < 0 , 0 p 7do2 <0, Sf3/5k_bar>0 (7) The result (7) is shown to explain the relationship between the government's optimal incentive scheme and the firms' performances. When the Balance of Payments Debt increases, the government changes the optimal incentive scheme in a way of increasing together. When the international pressures increase, the government decreases the incentive parameter in a way that the firm sequentially responds to this parameter change. Thus, when the economic environments changed, we get the results of how the incentive affects the Government's decision and Firms' decision variables. This model is of originality in a sense that first, it introduced the moral hazard and dynamic games to explain the relationship between the reform and economic 12. U0 is the reservation utility 13. Suppose there are k constraints involved. The Lagrangian function is, k L = / ( x , , ......,*„)+ £ JLj g J(x .1»......’*») • j =1 The second order sufficient condition is derived with the Bordered Hessian Matrix, noting how the first partials o f the constraint function g border the cross-partials o f L with respect to x i, x„. The bordered Hessian o f this Lagrangian is; H = L , . . . ! nl nn g l - - - - g ' n g g n g g I 0 0 g gn .0 .0 For a maximum, the border-preserving principal minors o f order k > k alternate in sign, beginning with (-1 )k + 1 , the second o f opposite sign, etc. Silberberg & Suen (2000), The Structure o f Economics A Mathematical Analysis, 117-150 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 13 performance in a political economy model of heterogeneity of interests between the firm and the government. Second, it introduced the risk, international pressures, wage rate, and BOP debt constraint. So when these parameters changes, we can see how the choice variables (optimal incentive parameter, and Firms high input or low input) changes responding. Third, it used the backward-induction way to explain the firms' behavior. Fourth, it assumed that the economy is the small open economy which will accept the price. II. Political Economy of Export Support Financial Policy (ESFP)1 4 - Centered on pre-structural adjustment and post-structural adjustment Throughout this chapter, with the consideration of the economic model I presented at chapter I, it will be descriptively explained that how the Export Support Financial Policy has been changing interactively with the international political economy factor and how this change affects firms' commitment to exports. And also, we will look at how this export commitment affects the government ESFP loan commitment over the pre- and post-structural adjustment. To understand the economic model presented here, we will focus on international political economy, the relationships of government and firm, the R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 14 relationships of government and financial markets, changes of export support financial policy, and its characteristics. For the case study and empirical analysis, I use the experiences of South Korea's economic development. I shall first present the concepts of policy loan, Export Support Loan Policies, and International Political Economy factors in two different periods over four different development policy stages. We investigate how export support loan policy changes depending on different periods and how international political economy places its presence on these policies. The data for policy changes, export changes and interest rates will be used for the empirical analysis to support the model's argument. A. Definitions of Policy Loan 1. What is policy loan? Policy loans would be one of the policy instruments of the government to supply subsidy-type loans to the specific industry. These kinds of policy loans have been existed in different forms and contents depending on the economic development and financial market conditions over many countries in the world. For example South Korea through 70s, 80s and 90s before the free market interest rate, due to the government regulated lower interest rate than market interest rate, there used to be a chronic excess money demand over decades of years. Almost all 14. The very detailed version of the second part is available upon the request. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 15 of the loans from the official financial system would be estimated to include the subsidy-type loans due to the fact of ceiled lower interest rate regulated by the government. There are two ways to define the Policy Loans. One is narrow way and the other is comprehensive way1 5 . Comprehensive policy loan has been treated by the various institutions and has been in various changes over the decades of years. So it's hard to get the consistent data over the time series. Therefore, we use the narrow meaning of policy loans to discuss about the Government Policy Loan systems. For the amount of policy loans, official statistics of the banks' loans are usually used. The definition of policy loans is the loans that government gives preferential amount of loans, conditions and interest rate1 6 to some industry in order to achieve a certain economic policy. The data used is official bank loans in the accounts of deposit banks according to policy loans1 7 . For the comprehensive way, it includes deposit banks' commercial bills discounted, directive financing1 8 , loans of export-import bank in addition to the narrow meaning of policy loans. 2. Changes of Policy Loans 15. KIIET, 1994,pl36-137 16. it includes lower interest rates, preferential loan conditions. Lee, ki young, “Policy loan’s effectiveness and improvement”, 1994, Korea Tax Institute 17. through the banking system, there are agriculture/fishery/livestock loans, rural private debt substitution loans, trade loans, loans for equipment of export industry, loans for energy conservations, housing loans, loans with government funds, and loans with National Investment Funds. Research Report 311, KIET, 1994, p. 136. 18. a directive financing is for example, Required loan ratio for SMEs. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. This kind of policy loan has been justified to cope with the market failures which failed in distributing economic resources efficiently and income through markets. Through 1960s and 70s, South Korea justified this policy loans to promote the exports and raise the heavy & chemical industry as a strategic industry. < Table 2-l> Objectives o : Policy Loans Period Policy Objective 1960s, 1970s Fast Economic Growth: Industrial Economic Policy 1. Export support 2. Heavy Industry Since 1980s Correction of Market Failure: to Balanced Economic Growth 1. SME 2. Agriculture/fishery/livestock 3. Housing sectors Relatively increases to the portion of industrial economic policy Sources: Economic Planning Board (EPB), Economic Report, each year Since 1980s, as results of unbalanced growth policies in 1960s and 1970s, market failures in Small & Medium Size Enterprises, Housing sectors, and Agriculture sectors were emerging. At table 2-2, correcting-market failure loans were 25.9% in 1980 and increased up to 43.6% in 1992 in total policy loans1 9 . The share of Trade loans and loans for export equipment decreased from 19.9% in 1980 to 8.6 % in 1992 while SME loans increased from 4.7% in 1980 to 6.9 in 1992. < Table 2-2> the trend of two policy loans to total policy loans (%) 1970 1980 1985 1988 1990 1992 1 9 . Total Policy loans at Table 2-2: Industrial policy loan + correcting-market failure loans. Source: Bank of Korea, each year. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 17 Industrial Policy Loans 59.1 74.2 72.1 68.5 62.4 56.4 - Development Institution 34.3 31.7 28.3 24.1 22.8 22.0 loans2 0 Trade loans 14.8 19.6 15.0 3.7 4.3 3.9 Loans for equipment of 0.0 0.3 2.9 8.5 6.6 4.7 export industry National Investment 0.0 4.6 4.6 3.3 2.2 1.2 Fund 10.0 8.5 12.8 13.9 14.4 17.1 Commercial Bills Discounted 0.0 9.5 8.5 15.0 12.1 7.5 others Correcting-market failure 40.9 25.9 27.9 31.5 37.5 43.6 Loans 3.9 4.7 3.8 3.6 6.4 6.9 SME loans 11.8 4.8 5.3 7.4 15.7 17.1 Agriculture/fishery/live 5.8 10.7 11.2 12.8 8.2 8.5 stock Housing loans Total Loans 100 100 100 100 100 100 Sources: Bank of Korea, 'Economic Statistics Yearbook', each year. Table 2-2 continued From the trend of Policy loan <table 2-2-l>, the growth rate of policy loan in the deposit banks decreased in 1987 and 1988 and it came back to over 20 % in 1989 and 1990. After 1991, it showed a trend of decrease. <Table 2-2~l> the trend of the amount of policy loan in the Deposit Money Banks, unit: 100 mill won. Policy Loans 1986 1987 1988 1989 1990 1991 1992 Agricult ure loans 5,949 7,231 12,020 14,176 19,192 23,310 26,145 0.05 0.05 0.08 0.08 0.09 0.09 0.09 Fishery loans 2,662 2,899 3,395 3,920 4,749 5,229 5,076 0.02 0.02 0.02 0.02 0.02 0.02 0.02 Livestoc k loans 3,272 3,207 3,499 3,813 4,848 6,530 8,761 0.03 0.02 0.02 0.02 0.02 0.03 0.03 - 4,880 4,742 4,459 4,304 4,104 3,914 20. Development Institution Loans: Industrial Banks’ loans and Export & Import Banks loans R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 18 Substitut ion loans of farms' and fishery's debts - 4,880 4,742 4,459 4,304 4,104 3,914 0.04 0.03 0.03 0.02 0.02 0.01 Trade Loans 34,445 24,204 12,016 13,822 19,473 22,543 25,422 0.28 0.18 0.08 0.08 0.09 0.09 0.09 Loans for export industry equipme nt 18,669 24,165 27,258 29,050 30,115 32,011 30,439 0.15 0.18 0.19 0.17 0.14 0.13 0.11 Energy saving loans 2,278 2,780 2,282 3,311 3,022 285 137 0.02 0.02 0.02 0.02 0.01 0.001 0.001 Housing Loans 26,083 30,204 44,465 60,157 78,060 97,078 119,95 5 0.21 0.23 0.31 0.34 0.37 0.39 0.42 Budget funds 18,871 21,394 24,672 31,478 37,469 46,206 55,683 0.15 0.16 0.17 0.18 0.18 0.19 0.2 National investme nt funds 10,550 10,671 10,761 10,533 10,238 9,839 8,033 0.09 0.1 0.08 0.06 0.05 0.04 0.03 Total policy loans (A) (Growth Rate, %) 123,229 (24.1) 131,635 (6.8) 145,710 (10.7) 175,279 (20.3) 211,470 (20.6) 247,135 (16.9) 283,56 5 (14.7) Total Loans(B) (Growth Rate, %) 390,986 (15.6) 430,958 (10.2) 488,054 (13.2) 625,478 (28.1) 740,286 (18.3) 894,156 (20.8) 1,027,9 70 (15.0) Policy loan share in total Loans: A/B (%) 31.5 30.5 29.9 28.0 28.6 27.6 2 7 .6 Share of ESFP loan in 43.1 36.7 27.0 24.5 23.4 22.1 19.7 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 19 total policy loans* Sources: Bank of Korea, Statistical Year Book, each year Comment: ( ), inside the table, shows the ratio in the policy loans in each year, "*" ; (Trade Loans+Loans for export industry equipment) / Total Policy Loan Table 2-2-1 continued The share of policy loan to total loans in deposit banks has shown a trend of decrease except 1993, but still has a substantial portion in the deposit banks' accounts2 1 . The reason that an increase of the share in 1993 is the result of the new policy called "Increasing the competitiveness of the Manufacturing Industry in 1993". Agriculture/fishery/livestock loans do not show a big change, but the housing loans increased from 21% to over 40% in 1993 in the total loans. The ESFP loan's share2 2 came down below 30 % in the total policy loans from 19882 3 and fell down to about 20 % in 1992. I mentioned the characteristics of the policy loans in the beginning that those policy loans have the subsidy-type loans under the Ceiled Interest Rate by the government. Therefore, decreasing the share of the policy loans in the total loans in the deposit banks and the share of the ESFP loan in the total policy loans tells us the decreasing of the subsidy-type loans to the firms who have been enjoying the lower interest rates and preferential loan amount. 21. Substantial portion means banks still has little autonomy in loan decisions. 22. The share of the ESFP loans at this table is the sum of the share of Trade Loans and Equipment for export industry. 23. In 1988, the government abolished the trade loans for large enterprises. Source: Bank of Korea, 1996 R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 2 0 From the above analyses, we could conclude there are characteristics of policy loans. 1) Except the ESFP loans, almost all of the policy loans were concentrated on the manufacturing industry, 2) the policy loans has been dealt with all of the banks throughout the whole financial systems, 3) since the target industry, the amount of loans, and the firms for the policy loans are decided by the policy decision of the government, the financial institutions didn't follow the assessment procedures of credit analysis, and profitability of the projects, but just follow the government policy, 4) in order to provide these loans at a lower ceiled interest rate, the government has to make up the difference of ceiled interest rate and market interest rate or the Central Bank had to provide the high powered money, 5) the policy loans were exempted from the credit regulation system2 4 and were enforced directly and indirectly by the government. 2 4 . Credit Regulation System is basically a system that reduces the credit concentration. It was a start of March 1969, however, it was from 1974 that actively “Credit Management Agreement to the affiliates” was enacted. Two things are basics. One is “Financial Institution Credit Management Regulations” and the other is “Credit Management Sections to the Affiliated companies”. CRS’ major contents are five ones. First, credit limit management about the affiliated companies. Second, firm investment and permit of owning real estates. Third, self-financing. Fourth, sales of non business real estates. Fifth, inducing business specialization. “Credit limit management about the affiliated companies” is to set the credit limit against top 30 chaebols’ outstanding loans. In case that equity/capital ratio is weakened conspicuously or the loan is increased to the very high level, this system let the companies sell their stocks and real estates to pay back the loans, source: ‘Institutional changes and economic performance on the Korean Chaebol’, 2000, Junkyu Lee R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 21 The ESFP has been positioned in the unique roles in leading the government policy which took guided measures to develop the export oriented development model. Until trade surplus in 1988, the ESFP is the most important policy in the export orientation and has the major share in the policy loans2 5 . Since a specific policy loan system could have multiple functions for the industry or various different loans would serve one goal, this whole policy loan system will necessarily bring the problems of transparency and efficiency. This kind of policy system would be followed by moral hazard problem for the government who guides the economy and the firm who follows the government's leading. Many firms in developing countries are used to this kind of Export Support Policy Loan given by their own governments. This Export Support Financial Policy has been changed over developmental stages with various other factors. In the following section, we will look how ESFP has been changed over different four developmental stages in Korea with other factors influencing ESFP. B. Definitions, Forms, and Procedures: Export Support Financial Policy (ESFP) In this section, I present definitions, formations, and procedures of Export Support Financial Policy. Since it was created in 1961, it has been modified a few times. It has the form of lower interest rates and amounts of loans as the benefits of this policy. 1. Definitions, Formations and Procedures R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 ESFP is a certain financial policy that supplies the loans to the firms which are involved in the exporting industries2 6 . In the narrow meaning of ESFP, it means that short term loans for the cost of exporting goods and loan guarantees related to the exporting. ESFP has been a major policy in South Korea in playing a big role in economic growth since it was created. (a) Foundations of ESFP ESFP was founded by the Financial Monetary Committee in February of 1961 to increase the exports with the 'Export Financing Ordinance'. Depending on the economic situations, it has been modified a few times. Current ESFP is taken as of March 15th in 1994 in the "Bank of Korea Total Loan Limit and Trade Loans Rules". (1) Contents of ESFP ESFP is into three areas in the Trade Financing. (a) General Export and Import Financing: In Trade Financing, related to the export and import, it supplies the loan before shipping goods to the exporter or export good manufacturer. It is for manufacturing and purchasing of export goods. (b) Construction/service Export Financing: it is the favored financing that is provided for the firms that earns foreign currencies by construction, shipping, and airline service businesses. 25. Sungsup Lee, “the rationalization of Trade Loans”, ‘National Assembly Bulletin 1992” p62-63. 26. It can include Trade Financing, Trade Bill System, Export Insurance System, Export Industry Equipment Financing, Deferred-Payment Export Financing, and Export Credit Guarantee System R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 23 (c) Agriculture/Fishery/livestock Export Preparation Financing: it is made to promote the agriculture, fishery, and livestock exports. It is providing the loans to stock and prepare the agriculture, fishery and livestock exports which are exported for the specific periods in mass amounts. 2. Sources of ESFP and Support Procedures The Sources of Export Support Financing Fund is made through types of discounting by the central bank. It is central bank discount rate-type policy loans. Therefore, sources of Export Support Financing Fund is made by the Central Bank and the supply of this fund is made through the banks which has the foreign exchange businesses2 7 . The supply channel is this. <Box 2-l> The Sources of ESFP and Supply Channel_____________________ Bank o f Korea — * • Discounting by central bank — > High Powered Money — ► Foreign Exchange Banks — > to the Exporting Firms. Sources: Ki Young Lee, ‘Policy Fund, Analysis o f Effect and Improvement’, Korea Tax Institute, p. 16 Therefore, the amount of ESFP played the role to influence the amount of high powered money. Because it is through Central Bank Discounting Box 2-2 shows the procedures of ESP. It shows that (i) 1s t stage: Export Contract is made, L/C arrives and Export Approval (ii) 2n d stage: ESFP loans (iii) 3rd stage: 2 7 Kyunghee Oh, ‘Explaining The Practice of Trade Loans’, 1986, p.20. R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 24 exporters' Shipping exports and banks' collecting the export loans that was loaned out. < Box 2-2> Export procedures and ESP supply procedures A Contract of Exporting — > L/C — ► Approval of Export — > ESFP: loans for Exports — > Shipping Exports — > Banks' collecting the Export Support loans 3. The Forms of Loan Procedures There were two forms of the loan process. The one is the L/C (Letters of Credit) base loan process and the other is actual export volume base process. All of the exporting firms were required to choose either one of them and they couldn't use both procedures at a same time. (a) L/C base loan process: Exporters should have to present the L/C to the banks and can receive the loans or payment guarantees based on the amount of L/C. Exporters are required to pay back the loans to the bank once the exports are paid. Since the bank estimates the amount of loan demanded based on L/C amount, the amount of loan is usually calculated in an accurate way. Yet, for the exporters, there are inconveniences of documentations every time they have to hand in L/C. Therefore there are shortcomings of administrative procedures like delays of approval. (b) Actual export volume base process R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 25 This is introduced to make up the shortcomings of L/C base loan process. It doesn't require the firms to hand in L/C. The bank makes a loan or payment guarantees within the limit of the trade loans based on export performances over a specified period of time in the past. The bank will collect the loans when the due date matures. By this method, the exporters borrow the trade loans even before the L/C arrives. This is very convenient process for firms. 80% of ESFP loan was done through this previous export volume record established. They can get the raw materials for exports, make the exports and ship earlier. This early shipment makes the loans paid off earlier as soon as the exports are paid. 4. ESFPs Policy Procedures The decision of ESFP is made by Financial Monetary Operation Committee (9 members) in the Central Bank. The policy alternatives are designed by the department of finance of Bank of Korea in the discussion with the department of financial policy in the Ministry of Finance2 8 . <Table 2-3 > Members of Financial Monetary Operation Committee in 1997______ Class Department The number of committee Chairman Minister of Ministry of Finance 1 The committee members The president of Bank of Korea 1 28. Bank of Korea, Financial Policy department, 1997 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 26 One member of committee recommended by Ministry of Finance 1 One member of committee recommended by Financial institutions 2 One member of committee recommended by Ministry of Trade & Industry 2 One member of committee recommended by Ministry of Agriculture & Forestry 2 Sources: Bank of Korea Law, Section 8th in 1997. Currently (2002) there is a change. The Chairman is the president of Bank of Korea. Bank of Korea becomes more independent from the government since 1998. Table 2-3 continued. Policy alternatives would be discussed in the discussion committee of Ministry of Trade & Industry, Ministry of Agriculture & Forestry with the Ministry of Finance. The ESFP is performed by the department of Finance in the Bank of Korea. 5. The forms of ESFP The ESFP has two forms of supports in terms of Ceiled Lower Interest Rate and Preferential Amount of Loans. (a) Interest Rate2 9 : Government applied the international level of interest rates to exporting firms and gave a preferentially lower interest rate compared to the domestic interest rate on general loans charged by commercial banks. Regulated export support interest rates helped decrease the production cost and increase the competitiveness of export goods in international markets. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 27 (b) Loans of amount Until 19943 0 , there were almost no restrictions for the amount of Export Support Financial Loans as long as the required qualifications are met. It is not subject to the Total Credit Regulation policy3 1 . Since the ESFP loans to the export industry were in the first priority in the export support policy system, it is provided beyond banks credit limits by the government and central bank. ESFP was simply outside the Credit Regulation system. But there was an important change in 1994. Since 1994, the bank of Korea took Total Limit Credit System and absorbed the trade loans into the general loan system. Due to this change, ESFP has changed from the direct support system into the indirect support system. Another important change is for the interest rate in 1992 and 1995. In 1992, the ceiled interest rates in the banking system were freed into market interest rates. Due to this influence on the ESFP loans, ESFP loans were provided at a bank prime rate level and even in 1995, ESFP loan interest rates were totally freed3 2 into market interest rate3 3 in 1995. 6. Background of introducing the Export Support Financial Policy 29. Ham, Sungho, ‘Understanding of Korean Financial System’, 1996, p300-306 30. Total Loans Limit System is imposed since 1994 to control the total credits in the economy. 31. In terms of Credit Regulation System, the total amount of domestic credit would be subject to that policy. . After November 1 in 1993, each financial institution can decide the interest rate for the loans freely. 33. This is a part in the three steps of Free Market Interest Rate policy in the Financial Reforms in 1990s. Source: Monetary Reform Committee, 1997, pi 10-118 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 28 • Korean War in 1950 and the economy was destructed • US Aid decreased a lot in 1960s, • Military government in 5.16 in 1961 • To get out of the absolute poverty, 5 year economic plan started • In Feb 1,1961: Financial Monetary committee; revision of Export Finance Code • In 1962: Revision of Bank of Korea Law; the authority of monetary policy is transferred to the hand of Ministry of Finance. With the high powered money of central bank, the government came to support ESFP loans in 1962. ESFP loans were extended to the exporters who received the L/C, export good manufacturers who received the order of the export products, and even producers who produced the raw materials for the exports3 4 . In 1965, as of the year Korea's entering the GATT, the direct export subsidy called "Export Encouragement Subsidy System" was not allowed to be used. Since then, the extension of ESFP loans has been emphasized more. C. Changes of ESFP with various factors: Pre- and Post-Structural Adjustment Periods In this section, I investigate the changes of ESFP, Economic environments interacting with ESFP decisions and International Political Economy Factors which influenced the domestic policy decisions over four different periods. 3 4 . Youngchun Kim, “a study of the effectiveness of Korean Trade Policy”, Kukmin Univ, 1992 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 9 This case study would tell us how favorably or unfavorably the government set the incentive structure where a firm is operating to maximize its benefits. In the first and second periods (1962-1979), the government guided strongly the economy and received little pressure from international financial institutions on ESFP. A firm enjoyed the great loan support from government in the export business. Once there were clear incentive systems for the exports, what firm does was to just follow the incentives set by the government. In the third and fourth periods (1980-1997), there are structural adjustments in the ESFP and firms no longer enjoy such great support from government. Once the incentive system turned unfavorable to firms, firms would face various other choices to maximize benefits. Under moral hazard, this firm may no longer choose export-driven maximum effort level which might be followed by upcoming macroeconomic turbulences. First Period: 1962-1971, Export Orientation centered on light industry Second period: 1972-1979, Export Priority Policy Period centered on heavy industry Third period: 1980-1987, Adjustment Period of Export Priority Policy Fourth Period: Since 1988, Openness and Globalization 1. Pre-Structural Adjustment (1962-1971): Export Orientation Policy emphasized on the Light Industry (a) International Political Economy: No pressure under GATT R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 30 During the two World Wars, U.S. emerged as the hegemonic state and assumed the major role in forming and maintaining the international systems such as IMF, IBRD, and GATT for Free Trade. The fundamental foreign economic policies of U.S. are the followings: (i) under GATT system, setting up free and non-preferential international trade systems with decreasing the tariffs to promote the free trade, (ii) supporting the developing countries' economic development with the favored and exceptional measures to promote, (iii) with the control of export of strategic materials to the communist regimes, restraining and retarding the economic development of the communist countries. Under this foreign economic policy, the U.S. enacted the Trade Extension Act (TEA) in 1962 to take the drastic decrease of tariff rate. This TEA contributed to the tariff decrease of the GATT system. U.S. economic policy for South Korea started with a great deal of economic aids. But the U.S. changed the aid policy from the grant-type free aid to credit assistance to decrease the U.S. budget deficit in the early 1960s. She proposed three plans as alternatives for the free aid to the South Korea: (i) changing the economic policy into the export priority, (ii) injecting the Japanese capital into the South Korea with recovering the diplomatic relationships between Korea and Japan, (iii) encouraging the U.S. firms' investment into the South Korea. Those suggestions were adopted by the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 31 South Korean government. The exports to U.S. were 400 million dollars in 1964 and increased to 530 million dollars in 1971 due to such policies. One of the significant reasons why the exports to U.S. increased a lot was because the urgency of South Korea's economic improvement was fit in the need of U.S. political economic situation after the World War Two against the communist regimes. U.S. needed desperately the political stability and economic development of South Korea as the geo-political importance to restrain the extension of the Soviet Communists since the U.S. set the first priority of foreign policy to curb the communism. Under the peak period of the Cold War, the U.S. provided the capitals, technology and markets for the South Korea as the part of the U.S. foreign policy strategy. She embraced the South Korea's exports with the strong economic power3 5 . Therefore, the U.S. government did not give any pressures on the international trade policy of the South Korea due to the U.S. foreign policy and small economy of Korea. During this period, the South Korea was able to adopt and take the trade policy aggressively without the international pressures. (b) A Relationships of Government and Firm: Strong Stackelberg Leader 3 5 . Kim Jung Su, “International External Effect of the policy and the dynamics of the trade conflict between the South Korea and the U.S.,” p, 180-181 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 32 In 1950s, the industrial policy3 6 was the import-substitution industrialization3 7 . The government budget was largely composed of U.S. foreign aids. In the late 1950s, when the U.S. economic aids decreased, production of the firms operating on the U.S. foreign aids decreased and unemployment rate increased3 8 . When the South Korea experienced the economic and social crisis in the early 1960s, the military general Park Jung Hee toppled the civilian government and became the South Korea's President as a military dictator. To fill the lack of legitimacy of political power and get Korea out of absolute poverty, President Park set off five year economic development plans3 9 since 1962. Since 1966, in order to increase the export, every month the president Park himself presided over the Export Promotion Meeting and Monthly Discussion Meeting with officials and business leaders. President Park placed the Export Promotions as the first priority in order to advance economic development4 0 . In 1967, South Korea entered into the GATT membership, enacted the Trade Transaction Act, and centralized the 36. Kim Suk Jun, “Economic Democratic Policy, State’s ability, the Role of the State”, State and Public Policy, Bub-Mun-Sa, 1991, p.481: (1) In 1960s: Export Orientation Industrialization Period, (2) In 1970s: Export Leading Industrialization Period, (3) In 1980s, Economic Liberalizaton Period, (4) in 1990s, Economic Democratization Period. In this paper, I will basically adopt his classifications of the periods. 37. Tun-jen Cheng, “Political Regimes and Developing Strategies,” in Gary Gereffi and Donald Wyman eds. Manufacturing Miracles, Princeton University Press, 1990, pl40. 38. the growth rate of GNP was 8.1% in 1957, 6.5% in 1958, and 2.5% in 1960. 39. the military promised the anti-communism and economic prosperity as a coup de’tat slogans. 4 0 . He seemed to believe that outward-oriented economies were better ale to gain. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 33 distribution of the domestic total credit and foreign borrowing loans to support the exports4 1 . Therefore, in this period, the government is a strong Stackelberg leader in leading the firms to respond the economic policies. For example, in the first five year economic development plan4 2 , each export policy goal was numbered in quantity of exports in detailed ways and the firms were responding to this number by ways of increasing exports to get rewarded by government. Once the government set up the industry development plan, the firms planned their own business plans corresponding to the government's plans and consulted with the government to reach the maximum goal. Once the firms' business plans were set in, the firms handed in the loan application forms to the government. The government determined the amount of financial support and approved the foreign loans to these firms who are willingness to run for the goals. A firm assumed the risk of business would be shared by the government who was willing to guarantee implicitly when things went wrong. Moral hazard situation would be a good fit in this kind of interactions between firm and government. 41. In 1965, Interest Rates Ceiling Law was enacted to maximize the mobilization of the domestic credit with realizing the interest rates. The government provided the foreign loans to export firms through the government loan guarantees together with the loans of nationalized banking institutions. 42. It started in 1962 and ended in 1966. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 34 The government set up the development plans. She rewarded the positively well-responding firms and penalized the badly-responding firms such as inclusions in the business, and loan supports. According to the government economic plans, the firms which were contributing to the economic development were given preferential financial supports in terms of foreign loans4 3 and domestic credits. Firms who received preferential loans and rates placed themselves in the much better position in businesses. The firms which were not responding well to the government economic development policy were penalized in terms of Internal Revenue Service Investigations and rejections of loan applications. (c) A Relationships of the Government and Financial System: Monetary Policy, BOK, Interest Rate Realization Act After the coup de'tat, the government first nationalized the banking system to control the financial market in 19614 4 . Nationalized banking system prepared the targeted distribution of the financial capitals for the specific industry where the government included in the economic development plans. The government permitted the foreign banks selectively in order to make up the shortage of the loans which was needed to support the 43. Government guaranteed the loans and provide the lower interest rates. 4 4 . Ku Suk Mo, “Growth and Stability Pendulum,” 40 year History of Korean Economic Policy, KFI, 1986, p. 127. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 35 development plans. In 1962, through the revision of Bank of Korea Act, the government came to have the monetary policy back in her hand. Due to this monetary policy transferred from the central bank to government, financing the fund for development became possible through high powered money. Foreign exchange policy and management were transferred to the government in the revised BOK Act. So the government prepared and utilized the monetary policy and foreign exchange policy for the economic development plans. Since then, government-leading economic growth system was established. In the early 1960s, the interest rates of the commercial banks were regulated by Financial Monetary Operation Committee and they were determined much lower than the market interest rates. To encourage the private savings to increase the domestic credit and balanced government budget, in 1965 the government enacted the Interest Rate Ceiling Act and increased the legal maximum interest rate up to 30% from 20%4 5 . But, this Interest Rate Realization Measure in 1965 enlarged the interest rate gap between the policy loan and general loan and therefore, the support for the export loans was enhanced by this 1965 measure. The government came to 45. Bank of Korea, “40 year history of Bank of Korea”, p.l 17. The intention of this 1965 Interest Rate Ceiling Act is to realize the regulated interest rate to the market rate, to increase the savings for financing the industrialization by government and to include the money into the banking systems. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 36 control more of the financial systems and the role of the government became stronger due to this Interest Rate Realization Measure. (d) The Analysis of the Export Support Financial Policies: 1962-71 (1) ESFP in the First and Second 5 year Economic Development Plans: 1962-71 Impressive Korean industrialization started from the first five year economic development plan (1962-1966). This first five year plan focused on establishing self- supporting economy from the Aid Dependent Economy. The second five year economic plan (67-71) focused on exporting products of light industry, developing the import substitution industry and incorporated the Korean Industry into the international division of labor. So at this second five year plan, the export orientation policy is the center of the economic policy4 6 . The growth rate was 2.2% in 1962 and since 1966, and it became over two digits. The exports grew at 40% average per year and it was 1.7 billion dollars in 1971. <Table 2-4> General Economic Data Contents 1962 1966 1969 1971 Average annual growth rate, % 1. The growth rate of the 2.2 12.7 13.8 4.6 8.3 economy, % 2. Industrial Structure, % - Agriculture/fishery 37.0 34.8 27.9 26.7 - manufacturing 14.4 18.6 20.3 22.2 3. Manufacturing structure - Light 71.4 65.9 62.4 63.0 - Heavy 28.6 34.1 37.6 37.0 46. Korea Export & Import Bank, ‘the 20 year history of Korea Export & Import Bank’, 1996, p.90 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 37 3. Export (million dollars) 55 250 658 1,677 40.4 4. Import (million dollars) 390 680 1,650 2,250 43.3 5. Trade Balances -56 -103 -549 -371 ~ Sources: Bank of Korea, Statistical Year Books, each year. Table 2-4 continued. At the table 2-5, in this period (1962-71), the Code of Export Financing was enacted (1961) for the first time in the government economic policy and the various measures of the Export Support Financing Policy were introduced with the increased amount of exports. The main frame of the L/C based financial support policy measures was introduced in this period, too. In 1962, the military goods suppliers4 7 were included in the beneficiary of this ESFP and in 1967, construction companies who were earning the dollars in foreign countries. This 1967's policy helped the construction enterprises compete in the foreign construction markets. In November 1969, the code of financing of agriculture/fishery export preparation was introduced to help increase the income of agriculture/fishery industry by exports4 8 . Especially in 1967, there was important enactment that extended the Export Support Financing to the producers who produced the raw materials which were used for the exports. It was only applied to the exporters but since 1967 it was extended to even the producers. This 1967's policy extended the beneficiaries of the 47. These suppliers provided the goods for UN army that were dispatched in South Korea at that time. 48. Usually agriculture/fishery industry export goods have a tendency to be stocked for some period of time before actual exports. This financial support helped financing the storage equipment and facilities. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 38 ESFP substantially and helped increasing the exports with the lower production cost4 9 . <Table 2-5> Changes of Export Support Financial Policy Year Major Policy Contents 1961 Enacting the code of Export Financing ESFP came to have the form of one of the financial system 1962 Enacting the code of military goods supplier earning dollars domestically Supporting the military goods supplier who are providing goods for UN army 1967 Enlarging the extent of the borrower Including Construction companies earning dollars in foreign countries into the beneficiaries of Export Support Financing 1967 Enacting the code of Import Financing 1967 Introducing the code of Domestic L/C system Supporting Domestic Enterprises which produce the raw materials involved in exports 1969 Enacting the code of financing of agriculture/fishery export preparation Enacting: in order to increase the income in the industry by increasing agriculture/fishery exports 1969 Enacting the code of financing of export raw material production costs Supporting the raw material producer for export goods Source: Department of Finance in Bank of Korea, 1996. 49. Because these ESFP loans have ceiled lower interest rate which was a kind of subsidy-type. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 39 At table 2-5, there were the status of ESFP loans and total loans in Bank Funds. The ESFP loans were 1.8 billion won, about 17 billion won in 1967, and 80 billion won in 1971 in table 1.3. The share of ESFP in total loans in the bank funds was 4.2% in 1962,10.9% in 1967 and 9.6% in 1971. In 1967, the growth rate of ESFP was over 200% due to the extending the beneficiaries of the Export Support Financing - construction companies, and raw material producers and domestic L/C system5 0 . Table 2-6> Status o ! ESF Loans (100 millions Won, %) in the Deposit Vloney Banks ESF: A Growth rate of ESFP: % Total Loans with Banking funds: B Growth Rate of the loans in the bank funds:% (A/B)*100 1962 18 - 432 - 4.2 1963 27 50.0 490 13.4 5.5 1964 25 -7.4 530 8.2 4.7 1965 46 84.0 564 6.4 8.2 1966 49 6.5 850 50.7 5.8 1967 167 240.8 1529 79.9 10.9 1968 245 46.7 2968 94.1 8.3 1969 351 43.3 5105 72.0 6.9 1970 559 59.3 6496 27.2 8.6 1971 801 43.3 8376 28.9 9.6 Sources: Bank of Korea, Monthly Economic Bulletin. The loan system of the unit price per dollar for Export Support Financing was introduced in 1964. So the exporters will borrow the amount of loans at this domestic currency rate per dollar based on the dollar amount export contracted. At 50. Domestic L/C is different from the Master L/C based on some conditions. 1) this is a kind of payment guarantees used for purchasing the imported raw materials for exports, domestically produced raw materials for exports or final products of exports. 2) this is only used for domestic transactions and prove the products was for exports in transactions domestically. Domestic L/C has benefits of earlier collection of sales, receipt of trade financing, zero value added tax, accepting the volume as established exports and duty return. Source: Bank of Korea, Trade loan procedures, 2001 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 0 the table 2-7, the domestic currency rate was 150 won per dollar in 1964,220 won per dollar in 1969 and 350 won in 19715 1 . <Table2-Z> Won per dollar for Export Financing, unit: Won Year Unit price per dollar for the loan contents 1962 Within the amount of L/C and other related costs 1964 150 Adopted the domestic currency unit per dollar for the loan 1965 200 1969 220 1970 240 1971 350 Source: Bank of Korea, Monthly Economic Bulletin, each month At the table 2-8, in 1962, ESFP loan rates was 9.1% and the spread with general loans was 7.3%. Export Support Financing Interest rate5 2 was down to 6.5% in 1965 and the general loans rate was 26% in 1965. The average spread of the two interest rates since 1965 was 18.2% compared to 7.3 before 1965. <Table 2-8> the spread of interest between ESF and General Loans5 3 , % ESF interest rate General loans rate Spread: B-A 1962 9.1 16.4 7.3 1965 6.5 26.0 19.5 1968 6.0 25.2 19.2 1970 6.0 24.0 18.0 1971 6.0 22.0 16.0 Sources: Bank of Korea, Economic Statistic Yearbook, each year. 51. Inflation rate between 1962 and 1971 was about 20% average per year and domestic currency rate per dollar for export support financing grew at about 25% average per year. 2 . ESF interest rate is the interest rate for the loans for exports of the Deposit Money Banks, Source: Bank of Korea, each year, Statistical Year Book 53. Call loans is the short-term overnight loans which is not fit in comparing with the export supporting loans which is usually over 3 months. Corporate bonds market was not developed well, too. Generally it is appropriate to use the loan rates on installment savings. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 41 Comments: 1. General loans rate: the interest rate of the loans on installment savings in the commercial banks. 2. ESF loans : loans for exports in the commercial banks. (2) Characteristics of ESFP in the First Period (1962-1971) In this period, for the international political economy, there was little pressure on the South Korea's economic policy due to the Cold War and Korea's small economy portion. The government took the initiative of Export Financial Policy with the strong state leadership. First, the government introduced the codes of Export Financing Loan Policy in 1961 and later took the various measures - financial systems for loans for production of raw materials for exports, loans for supplies in foreign currency, loan procedure codes for agriculture/fishery export preparations. Second, the government provides by far lower interest rate which gave firms financial benefits. ESFP loans were averagely 13% lower than the general loan interest rates. Third, the share of the ESF loans was 8.4% average per year relative to the total deposit money banks loans. Fourth, virtually there was no limit for the ESFP loans as long as the exporters met the qualifications5 4 because the loans were provided by the high powered money. <Table 2-9> The features of ESFP in 1962-1971 Supports Characteristics status Measures of Export Support Export Financing Code was founded and introduced in 1961. Production fund for export raw 5 4 . L/C and the established amount of exports. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 42 materials Construction & Service: beneficiary Domestic L/C Interest rate Export Financing loan interest rate was 7% average per year and general loan interest rate was 22% average per year the spread of interest rate was 15% subsidy-type loan Amount of ESF loans Total bank funds were 273.4 billion won average per year and ESF loans were 22.9 billion won average per year the share of ESF loan relative to the total bank funds: 8.4% Sources: Bank of Korea. Table 2-9 continued. 2. Pre-Structural Adjustment(1972-1979): Export Priority Policy Period (a) International Political Economy: Low Pressure Since 1970s, international economic horizon has been changed fast. U.S. absolute hegemony started to be weakening due to the EC and Japan's fast economic development. International free trade system which was based on the hegemony of U.S. could not exist any more. In the end, fixed exchange rate system which was based on the U.S. dollars had to be changed into the flexible exchange rate system. Due to the two oil shocks and stagflations worldwide, free trade system yielded to the neo-protectionism5 5 . This neo-protectionism is related to the weakening of the GATT system because the preferential measures were against the 55. Neo-protectionism: it is different from the 19th century classical protectionism. 1) the advanced countries took this protectionism, not the under-developing countries. 2) the protected industry is not the infant industry, but the industry which came to lose the competitiveness due to the delay of industry structural adjustment. 3) in the name of “market disturbance prevention”, trade measure was done in a way of preferential and targeted area. 4) the tools of protection are not the tariff but the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 43 multilateralism and non-conditional MFN (most favored nations) treatment which were bases of GATT system. Because of the increasing non-tariff barriers, GATT initiated the multi-lateral trade agreement in order to abolish the trade barrier. From 1973 to 1979, it was the Tokyo Round which initiated new Round. At the Tokyo Round, among the non tariff barrier areas, the most important subjects were the subsidy and Countervailing Duty Agreement5 6 . One of the characteristics in the Round was to allow the countervailing duty to be imposed if there is any violation of the subsidy stipulations. The change in the GATT system was due to politically the change in the U.S. hegemony caused by other emerging countries and due to economically the trade deficit of U.S. in itself. In 1971, since the WWII, U.S. experienced the first trade deficit about 2.3 billion dollars and after 1976, it became a chronic problem since then. In 1974, Trade Act was enacted and "Fair Trade" concept was introduced5 7 by the U.S. Congress. U.S. domestic industry strongly asked the US government to non-tariff measures like OMA (Orderly Market Arrangements) and VER (Voluntary Export Restraints), Kim Sewon, ‘International Economic Order’, 1987, p57-61 5 6 . Examples of the export support financing related subsidies on the subsidy and countervailing duty agreement presented by the GATT Section Contents K section • When the government provide the lower interest rate of the export support loans or the government pays a part of the cost in order for the exporters to finance the export loans Sources: Kim Jundong, ‘Surveys of export support loans of major countries and improvement plan for Korea’s system’,. KIEP, 1993 5 7 . “Fair Trade” concept was half withdrawn from the Free Trade and half forward from “Protectionism”. By Section 301 in Trade Act, the retaliation was allowed. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 4 protect its industry, too. By this Trade Act, U.S. started to regulate the free trade with the so called "Unfair Trade" countries. To South Korea, U.S. used the quota system, anti-dumping and countervailing duty, instead of applying 1974 Trade Act for the unfair trade. Therefore, the pressure to Korea was not so high and the international pressure on the Korea's targeted industrial economic policy decision was little. Throughout the 1970s, even though there were a lot of changes in the international economic systems, the influences on the export support financial policy of South Korea were not so great that the Korean government was able to keep on providing the intensive and extensive export support to the export industry firms. (b) A Relationships of Government and Firm: Strong Stackelberg Leader The Export Industry Priority System was simultaneously determined with the authoritarian regime since military coup d'etat. Both supported themselves each other strongly. The authoritarian government5 8 led the firms strongly into the export priority businesses and provided the full support for this drive. In order for the government to lead and support the firms, the government gave these full supports such as the August 3rd Policy, Trading Co, Heavy Industry supports with financial resources and tax treatments. Especially among those measures, the 58. “Yushin Cheje”. South Korean Authoritarian regime which started new Yushin Constitution that fortified the power of the president over the Parliament and the Justice. It gave a serious setback to the Korean democracy in 1972. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 45 August 3rd measure and the support for the Heavy Industry revealed close relationships of government and firm very well. The August 3rd Policy5 9 in 1973 was the unprecedented financial policy taken by the military government to save and protect the financially troubled firms. The government intervened financial market mechanism of the credits only to help the firms at the cost of most of other economic participants. This incident definitely showed the government7 s increase in strong control of the economy and in the intervention of the economy. The second support was to promote building the Trading Companies. In 1973, the government introduced the trading company systems in order to boost the volume of the exports. She provided the full supports and incentives6 0 to the trading companies. The government chose the 6 strategic Heavy industries - steel, chemical, nonferrous metal, machinery, shipbuilding, and electronics. Focused on six heavy industries, the government increased the supports financially and monetarily. In 59. In August 3r d , 1972, the Korean government took the very extreme financial interventions in the usurious private money market to help the firms. Major contents in the policy were to boost the firms’ financial status - clearing the private money market debts, lowering the interest rate and providing the extra funds. Due to this Measure, the benefits of the firms were assumed to be about 102.8 billion won. This was one of the examples where the Korean government intervened the market to support the firms in an unprecedented way. 6 0 . The supports and incentives for trading companies are followings: (1) Based on the export results, the rights of lower interest rates were given to the larger-export-volume firms. (2) Trading companies were given rights to borrow a loan from the foreign banks in the foreign countries. (3) The priority rights was given in international biddings. (4) Trading companies were given the first priority rights to set up the construction companies for the Middle Eastern Countries. (5) Trading companies were exempted the corporate tax up to 24%. (6) Trading companies can delay the payment of the business R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 46 order to increase the financing of the equipments in the heavy industry, on December in 1973, she established the 'National Investment Funds (NIF)' and had the Ministry of Finance manage and control this fund to support the heavy industry. In addition to this NIF, the deposits of the Korea Development Bank (KDB), and export industry equipment financing was preferentially given to the heavy industries. The share of the policy loans relative to the total loans in the banks was about 40% in the late 1960s and increased up to about 60% in the late 1970s. In 1974, through the Tax Exemption Regulation Act, the government gave tax exemption up to 100%6 1 for these heavy industries. Compared to the 1960s' industrial policy which were complacent to the comparative advantage, 1970s' heavy industry supports were determined mostly by non-economic decision.6 2 Therefore, the government led the whole industry into one direction of exports and heavy industries to increase the exports and self-reliant defense. The Korean government was a strong leader and the firms were followers to this government leadership. (c) Government and Financial System tax for the retained earnings in order to develop the export market. Suh Jaejin, ‘The Capitalist in South Korea’, 1991, p.97 61. (1) The first three years were 100% and next two years were 50%. (2) investment Tax Credit was 8-10%, (3) Special Depreciation was 100%. The firms chose one among three benefits. KDI, ‘The improvement plan of Industrial Policy and Supporting measures, 1982. 6 2 . Throughout the 1970s, Self-reliant Defense concept was first considered by then-President Park due to the movement of the withdrawal of U.S. soldiers stationed in South Korea. The Heavy industry requires the economies of the scale so that it requires the enormous amounts of investment. This asked the government to intervene the financial markets and to lead the firms strongly. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 47 Since 1960s, due to the fast economic growth and strong money demand, South Korean firms had been chronically lack of the loans what they needed. The financial system was forced to expand the size of the market instead of the structural improvement. As the economy grew fast, the financial system came to lose the direct role and to have more financial intermediaries. The financial market was relatively retarded and the financial intermediaries - i.e. banks - got bigger role in supplying loanable funds. The chronic excess demand for the money resulted in the firms' heavy dependence on the usurious private money market. In the early 1970s, m any firms found themselves in financial troubles because of the increase of inventory, the cost of manufacturing, financial cost and the burden of the foreign loans. Mixed with the dependence on the usurious private money market, the firms' financial cost became major burdens to many financially-already- troubled firms. The government was concerned about large increase in firms' financial costs and plunge-down of the credit rating in the international market for the firms. In order to save most of the financially troubled firms, encourage the savings and incorporate the private money markets into financial institutions, the government took extreme economic measures so called "August 3rd Policy". The major contents were: (1) In order to incorporate the private money markets into the financial institutions, the government enacted and announced the three Private Money Market Acts such as 'Short Term Financing A cf, 'M utual Savings & R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 48 Finance Acf, and 'Credit Union Act\ Based on these Acts, non-banking institutions increased up to 20 Short-term Financing companies, and 148 Credit Unions as of 1979, (2) Nullifying all private market's credits and debts and reporting all debts to the government, (3) Rescheduling of equally divided repayment of the only reported debts, (4) Transferring the short term high interest rate bank loans into the long term low interest rate bank loans, (5) Decreasing all of the interest rates, for example, general loan rates from 19% to 15.5% by government decree. All of the reported private loans was estimated about 80 percent of the money stock. By this August 3rd policy, the firms had tremendous benefits of as same as that amount of frozen private loan directly. This extreme government intervention caused the shocking wave over the Korean economy. This shock showed a comeback of the government intervening financial market. Since the August 3rd in 1972, most of these private money markets were incorporated into the non-banking institutions6 3 , and the government was able to control even the non-banking institutions with all other banking institutions. On December 1977, the government revised the "Bank of Korea Act" and added the vice-president of the Banking Supervision Authority as one of the board members in the Bank of Korea6 4 . It enacted the comprehensive order rights of the 63. So called “the Second Financial Institutions”. 64. It means that the government came to influence the decision-making of Bank of Korea. R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 49 chairman of the Banking Supervision Authority to influence the relationships between the banks and firms6 5 . Since 1972, the share of the bank loans in the total financing has been decreasing and the non-banking loans and direct financing has been increasing. At the table 2- 10 below, another share change for the foreign financing loan was due to the policy shift of the government. The government had been seeking more Foreign Direct Investment and less foreign loans in 1970s. <Table 2-10> Sources of financing Unit: % Contents 1970 1975 1980 Direct Financing 15.1 26.1 22.9 state bonds 0.1 0.8 0.9 - CP 0.0 1.6 5.0 Corporate 1.1 1.1 6.1 bonds Stocks 13.9 22.6 10.9 Indirect Financing 39.7 27.7 36.0 Banking 30.2 19.1 20.8 Loans Non 9.5 8.6 15.2 banking loans Foreign Loans 29.6 29.8 16.6 Others 15.6 16.4 24.5 Total 100 100 100 Sources: Bank of Korea, Financial Statement Analysis, each year. Table 2-10 continued Overall, the government strongly controlled the financial sectors and set the interest rates. It p u t the banking institutions and non-banking institutions under the control since the August 3rd measure in 1972 and intervened the whole financial process for the interest rates and the amounts of loans. We may call this period as R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 50 the extreme intervention period of the government in the financial institutions to support the export priority and heavy industries. (d) Export Support Financial System: 1972-1979 The South Korean government put the first priority of industrialization policy for the heavy and chemical industry during this period. The direction of the economic development was to develop the H&C industry into the export industry. The government picked machinery, steel, electronics and ship-buildings industry as infant industries and concentrated the investments on these industries helped the companies run for these businesses. During this period, at the table 2-11, the average economic growth rates were more than 9% and the manufacturing industry share surpassed the agricultural industry in the total industry structure. H&C became bigger than the light industry in the manufacturing structure. The exports were about 10 billion in 1977 and the average growth rate of exports was 29.3%. <Table 2-11> General Economic Data Contents 1973 1975 1977 1979 Average annual growth rate, % 1. The growth rate of the economy, % 2. Industrial Structure, % 12.6 6.1 10.1 6.8 9.3 - Agriculture/fishery 24.9 24.9 22.3 19.0 - manufacturing 3. Manufacturing structure 24.9 25.9 26.8 27.5 - Light 59.1 52.1 48.5 45.1 65. Choi Jin Bae,’Financial Policy after 1945’, 1996, pp94-95 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 51 - Heavy 40.9 47.9 51.5 54.9 3. Export (million dollars) 3,271 5,003 10,047 14,704 29.3 4. Import (million dollars) 3,837 6,674 10,523 19,100 32.0 5. Trade Balances -309 -1,887 -476 -4,151 ” Sources: Bank of Korea, Statistical Year Books, each year. Table 2-11 continued. The export support financing system was readjusted to meet the demand of the export priority economic policy and the focus on the heavy industry. In terms of policy operations of export support financing system, there came a simplification of the related export financing stipulations, swift operational responses to the export environments, and the increases of medium & long-term deferred payment exports. In terms of simplification of the stipulations, on March in 1972, the government consolidated the complicated various export support financing stipulations into the "Export Financing Stipulation". The export support financing system came to have just three branches stipulations after the overhaul since 1972. Those are 'Export Financing Stipulation', 'Loans for Supplies in Foreign currency Stipulation', and 'Agriculture-marine product Export Preparation Financing Stipulation'. In the area of the elastic responses to the export environments, the government took thoughtful consideration of the export financing systems to realistically help the export firms. For example, whenever there were important changes in the economic environment such as oil shocks, the government promptly adjusted the loan subjects, loan periods, loan rates and the loan procedures to meet the actual demands. Table 2-12 shows the major policy changes during this period. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 52 <Table 2-12> Changes of Export Support Financial Policy Year Major Policy Note March 1972 1. Merging four different stipulations into one stipulation: "Export Financing Stipulation" - export financing stipulation - pre-export financing stipulation - import financing stipulation - domestic material financing stipulation 2. maintaining the stipulation of loans for supplies in foreign currency 3. maintaining the stipulation of agriculture-marine product export preparation financing 1. Merged complex export supporting financing stipulations into the simple three ones. 2. this system was adjusted elastically to the economic situations such as oil shocks - adjust: borrowing firms, loan rates, loan periods, procedures. - delicate care of export support financing Sources: Department of Finance, Bank of Korea, 1996 Export Support Financing loans in the deposit money banks was 1084 million won in 1972 and increased more than 10 times larger in 1979 compared to the total loans' seven folded increases. The average share of the ESF loans to total bank loans was about 13.7% throughout this second period. Table 2-13 showed the status of ESF loans and total loans in the banking funds. <Table 2-13> Status of ESF Loans (100 millions Won, %) ESF: A Total Loans with Banking funds: B (A/B)*100 1972 1,084 11,024 9.8 1973 2,241 14,835 15.1 1974 3,602 22,798 15.8 1975 3.392 27,138 12.4 1976 4,618 34,481 13.4 1977 5.674 42,907 13.2 1978 8,832 60,191 14.7 1979 12,272 82,676 14.8 Sources: Bank of Korea, Monthly Economic Bulletin Comments: 1. status of ESF Loans is the loans for export supports in the deposit money banks. 2. B: Loans with banking funds. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 53 Export financing unit price had been 380 won per dollar in 1974,425 won in 1978, and up to 455 won per dollar in 1979. The unit price gradually increased almost every year. The loans for exports data was from the table of major interest rates on loans and discounts of deposit money banks in the BOK statistical year book. Actually ESF loan rates at table 2-14 included the rate for the loans for exports, the rate for the loans for supplies in foreign currency and the rate of the loans for export preparation of agriculture-marine products. <Table 2-14> Won per dollar for Export Financing Loans Unit: Won Year Unit price per dollar for the loan Contents 1974 380 1977 425 Gradual increase 1978 425 1979 455 Source: Bank of Korea, Monthly Economic Bulletin, each month. In the table 2-15, the loan rates of the export support financing was 6.0% in 1972, 9.0% in 1974, and 8.0% in 1978. The average ESF Loan rate was 7.75% over the period. The spread between the ESF Loan rate and general loan rate was averagely 9.1% over the period of 1972-79. The interest rate on loans for raw material imports was 9%. The ESF loan rate was roughly a half of the general loan rates, so getting the ESF loans was large benefits for the firms such as subsidy effects. <Table 2-15> the spread of interest rates between ESF and General Loans, % ESF interest rate: A General loans rate: B Spread: B-A 1972 6.0 16.5 10.5 1973 6.0 15.5 9.5 1974 9.0 15.5 6.5 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 54 1975 7.0 15.5 8.5 1976 8.0 17.0 9.0 1977 8.0 17.0 9.0 1978 9.0 19.0 10.0 1979 9.0 18.5 9.5 Sources: Bank of Korea, Economic Statistic Yearbook, each year. Comments: 1. ESF Loans: the interest rates on loans for exports of Deposit Money banks, p.22, statistical year book. Table 2-15 continued. (2) The Characteristics of ESFP Loans in the second period (1972-1979): There were little pressures from the international political economy factors. The South Korean government was able to determine the economic policy decision independently from the foreign countries and international financial institutions. The government-firm relationship was the government-led feature. For the domestic financial systems, the government strongly guided and controlled. Like the first period (1962-1971), fortified state-led economic systems were maintained throughout the period. The fourth republic adopted the Export Support Financial Policy without being intervened by the foreign countries' economic policy. (i) The government efficiently determined the support conditions, level of ESFP loan interest rates, and the size of the loans with the simplified export support systems. (ii) The second period export loan rates was averagely about 8.0% slightly higher than the first period rate was about 7%, but in terms of real R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 55 interest rate, the second period rate was smaller than the first period6 6 . The subsidy-type interest rate was lower in the second period in the real terms. (iii) In the volume of the ESF Loans, the average share of ESF Loans in the banking funds was 14% higher than 8.4% share of ESF Loans in the first period. In the second period, the government guided the firms, and controlled the financial market to give more preferential benefits to the export industry more than the first period. The table 2-16 gives the outlines of the ESFP loan policies in 1972- 79. <Table 2-16> the Features of ESFP ^oans in 1972-1979 Contents Features Supporting system - Simplifying and merging all littered supporting systems into the three areas - Effectively supporting the export industry Interest rates - average general loan rate was 17%, but ESF Loans was about 8%. 9% lower The amounts of ESFP Loans 8.4% share of ESF loans increased to 14% share of ESF loans in the banking funds in the second period. 3. Structural Adjustment Period (1980-1987): Adjustment of Export Policy and Reform-initiation (a) International Political Economy: high pressure 66. Inflation rate was averagely 15% in the first period and 21% in the second period. Bank of Korea, Economic Statistics Year Book, each year. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 56 In 1984, U.S. trade deficit was widened to 100 billion dollars and in 1985, U.S. became the net-debtor nation for the first time since the WWI. With the waning economic power of U.S., new emerging change came from the NICs 6 7 (Newly Industrialized Countries), Japan and the European countries. These NICs export volumes were closely following Japan's size in exports. Japanese economy emerged as a significant figure in the international economy with the European Community. The super powers in the international economy came to be shared by U.S., Japan and EC as the 1980s were unfolded. Two-layered International division of labor became three-layered division of labor - advanced countries, NICs and the developing countries. You see the changes of the share in the international trade for U.S., Japan, EC and NICs at the table 2-17 below. <Table 2-17> New Multiplolarized World Export Status________ Unit: Percentage 1960 1970 1980 1985 1990 World Export 100 100 100 100 100 U.S. 17.2 14.9 11.7 12.4 11.8 Japan 3.4 6.6 6.9 9.7 8.6 EC 27.1 29.2 26.3 25.6 29.5 NICs 2.5 2.5 13.7 6.3 8.0 Sources: Korea National Statistics Office, 'International DB', each year Comment: (1) EC: German, England, France, Italy. (2) NICs: South Korea, Taiwan, Singapore, Hong Kong. GATT in itself revealed the weakness & limitations to deal with widespread protectionism and managed-trade system taken by each country in 1980s. The limitations of GATT resulted from the vague stipulations, no effective punishment R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 57 system for the deviated states, and the structural change in the international political economy. In order to overcome these limitations, the world with the leading roles of U.S. and Japan started a new multilateral trade agreement so called 'Uruguay Round' on May in 1983. The 5th South Korean Republic6 8 was relatively more vulnerable6 9 in terms of the independent ability of international economic policy decision to the pressures by the U.S. government and international institutions. By the 1979 Trade Agreement Act, U.S. was able to retaliate the unfair trade countries with the countervailing duty against the subsidy, not only suing these countries to GATT. In 1984 Trade & Customs Act, the U.S. introduced the new concept of subsidy which considered even supplying benefits for raw materials, and intermediate goods as subsidies in addition to the subsidy for the final product. Since 1979 Trade Agreement Act, U.S. imposed the countervailing duty7 0 for the products which were alleged to be made with various forms of help of the export 67. South Korea, Singapore, Taiwan and Hong Kong are the NICs’ memberships. 68. President Jun Doo Whan’s government was called ‘the Fifth Republic’. 69. The first reason was President Jun did not have a democratic legitimacy. He took the power with a coup detat on May in 1980. The second reason was the Korean economy itself that had trade surplus with U.S. and bigger size of the economy in the international trade. The third reason was the unfavorably high interest rate in the international financial market. In the early 1980s, South Korea was in the foreign currency crisis situation. 70. <table > U.S. countervailing duty against South Korea’s products_____________________ Date Products Result Jan 19, 1982 Certain steel wire nails Cleared from the subsidy charge May 13, 1982 Large diameter and small diameter welded carbon steel pipes and tubes Withdrawal of the countervailing duty order May 19, 1982 Hot-rolled carbon steel plate Withdrawal of the countervailing duty order R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 58 support financing loans. Especially U.S. considered the Korea's export support financing as the critical subsidy policy and regarded 5-12% interest rate spreads between the market interest rate and the ESF loan rate as a subsidy. U.S. interpretation of the Korea's ESF loan as the subsidy was not specified even in the U.S. countervailing duty related Acts. U.S. extensively interpreted Tokyo Round's subsidy stipulations to apply the countervailing duty to Korean exports. U.S. pressured and made the Korean government open domestic goods market, financial market and stock market. U.S. also used the Super 301 Stipulation in the Trade Act to make the Korean government open fire insurance, marine insurance, shoes, steel products, and intellectual property rights market. Therefore, the South Korean government was not as much strong for the international economy policy decision-making as it was in the 1960s and 1970s. Especially, Korean government was weak to the U.S. pressure to open market. May 19, 1982 Hot-rolled carbon steel sheet Withdrawal of the countervailing duty order May 19, 1982 Galvanized carbon steel sheet Charged Jan 10, 1983 Bicycle tires and tubes Cleared from the subsidy charge June 13, 1984 Oil country tubular goods Cleared from the subsidy charge June 18, 1984 Cold-rolled carbon steel sheet Charged June 18, 1984 Cold-rolled carbon steel structural shapes Cleared from the subsidy charge Feb 11, 1985 Tapered tubular steel transmission Withdrawal of the complaint April 18, 1985 Offshore platform jackets and piles Charged Oct 24, 1986 Welded steel wire fabric products Withdrawal of the complaint June 30, 1988 Industrial fan belt Charged Sources: Korea Trade Association, each bulletin R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 59 (b) A Relationships of Government and Firms: For the Fifth Republic of South Korea, the relationship between the government and firms would be government dominant, but market prevailing in some ways. New military government since the assassination of President Park was aware of the government failures of the 1970s strong intervention in the financial markets. New military government wanted to decrease the government interventions and tried to make the private firms lead the economy. When General Chun took the power in 1980, he had to show he is committed to the new economic liberalizations to the international economic community, such as the World Bank and other international financial institutions to help the troubled economy in the early 1980s. New economic teams presented the policy visions of economic structural stabilization and economic liberalization to the new military leaders who did not have deep economic knowledge and understandings. These reformist technocrats pushed two major economic policies: (1) Economic Structural Stabilizations: heavy industry investment readjustment because the over-investment in this industry resulted in many defaulted firms. For the price stability, contractionary monetary policy was adopted to prevent the inflations7 1 . (2) For the Economic Liberalizations: 71. Since May 12 in 1973 when the government started the heavy industry support drive, the inflation rate increased up to about 31%. President Jun’s government understood this high inflation was the result of government’s over-intervention in the economy. The 1974-81 inflation rates were the followings. __________________________ _________________ _________________ _________ Year 74 75 76 77 78 79 80 81 Inflation rates, % 30.7 26.5 23.4 16.3 24.2 19.5 24.6 17.6 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 60 they propelled the import liberalizations7 2 , financial liberalization, and capital liberalization. Export priority policy was not on the list. On April in 1981, the government enacted the Fair Trade Act to promote the competition among firms and prevent the economic concentrations by the large enterprises7 3 . The Fifth Republic government seemed to focus on the less intervention and more competition-encouragement which was a good sign of economic liberalization. Firms just followed the government7 s new pro-competition direction in one side. However, for the Fifth Republic7 4 , the economic development was essential to get the approval of the Korean people's heart for President Chun. In contrary to the slogan of the economic liberalizations, in practice, the military government had to intervene in the goods and financial markets again to achieve the targeted economic development. This7 5 is the conflicting goal against the economic liberalization for the 72. Import liberalization meant the opening of the domestic market to the foreign products. Financial liberalization meant the banks’ privatization and self-management so that the decrease of the government intervention in the banks businesses was done. Capital market liberalization is to open the stock market to the foreigners. 73 . The concept “Conglomerates” was introduced at the FTC (fair trade commission) first revised law in 1986. FTC defined the companies that has over asset total o f400 billion Won as conglomerates and Limited Total Equity Investment (LTEI) & Banned on Mutual Equity Investment (BMEI)7 3 from 1986. Until now , the FTC has continued to introduce the Limit on Mutual Debt Guarantee (LMDG) and other policies in order to prevent the aggressive expansion of the conglomerates. 3ri, 4th , 5th and 6t h are more important due to introducing the new regulations. Topl-30 conglomerates concept in terms of the asset total were introduced at 1993 FTC ordinance. These asset total criteria top 30 conglomerates are currently applied so far. 74. Because the military general Jun took the power from the civilian government in 1980, it did not have the democratic legitimacy. 75. Economic Liberalizations and Interventions. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 61 Fifth Republic. The government sometimes gave the preferential business rights to some firms and encouraged the economic concentrations for the specific firms. One of the outstanding examples is "Government Forced M&A of the insolvent firms" in 1984-86. In the early 1980s, many firms which went to the Middle Eastern construction booms in 1970s became financially troubled due to the excessive business expansions and borrowings. The banks which guaranteed the debt of those firms by the government encouragement were troubled, either. In order to save these firms and the banks, the government, contrary to the early slogan "Economic Liberalizaton", strongly intervened in the M&A of these insolvent firms. The methods the government took are the third party acquisition 56 firms, consolidating the sister enterprises for 13 firms, liquidating 2 firms and work-out for 12 firms. During this process, the government introduced the exemption of the principal payment, indulgence of interest rate payment, new loans for the loss compensation, and deferment of the principal payment. It revived the 'Bank of Korea Special Loan Act' and passed the revised 'Tax Reduction Act* facing the strong resistance of the opposition party on December 1st in 1985. Since this kind of government intervention, the most benefited beneficiaries were the large firms. Among the insolvent firms, the 26 firms were handed to the 30 large enterprises. They enjoyed the immense financial benefits and acquired new firms easily with just jumping on the wagon of government policy. They were able to expand their size with the government supports. During the process of clearing R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 62 out the insolvent firms, lowered economic growth rates7 6 caused the government to increase in the government policy loan to cope with this lowered growth rates. 1982 1983 1984 1985 1986 1987 1988 Share of the policy loans in the total loans 39.7 40.6 40.7 40.1 40.5 38.7 37.6 Sources: Bank of Korea, Economic Statistics Yearbook, each year. Comment: total loans = total loans in the deposit money banks + loans in the Small & Medium Industry Bank + loans in the Export & Import Bank of Korea. Conclusively summarizing, the Fifth Republic showed government dominant economic system but market ruling some ways. The economic liberalization was initiated to solve the 4th Republic's failures and to respond to the pressures of U.S., not because the government was subject to the firms' economic leaderships. In the process of clearing out the insolvent firms, the government gave the firms strong guidance and heavily intervened in the market which often caused the economic concentrations for the large enterprises. (c) A Relationships of Government and Financial System The Fifth Republic promoted the Financial Liberalization in the early 1980s. The background of the financial liberalization was not only the domestic policy failures 1982 1983 1984 1985 1986 1987 1988 7.2 10.7 8.2 6.5 11.0 11.0 10.5 77. Here, the policy loan = narrow meaning of the policy loan + loans in the Small & Medium Industry bank + loans in the Export & Import Bank of Korea. It includes a larger meaning of the policy loan definition. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 63 of 1970s heavy industry support, but also the foreign pressures from U.S. and international financial institutions such as World Bank. U.S. firmly and steadily pressured the South Korean government to open the financial market and the World Bank set the precondition for the international loans as the financial market liberalization early in 1980 when the military government desperately needed the loans. In this period, the pressures of foreign institutions and states on export support policy were pronounced and put into place in the policy making. At next empirical section, we would see the influences of these foreign pressures on export support loans and exports with various empirical analyses. We expect the significance of those pressures on endogenous variables. The promoted financial liberalization7 8 involved encouragement of the self management of the banks, decreasing the entry barrier of the financial business, increasing the business areas of the financial institutions, interest-rate liberalization and capital market openings. First of all, for the self-management of the banks, the government promoted the privatization of the five big banks which resulted in all five banks privatization7 9 in 1983 since 1972. In order for the private usurious loan market to be incorporated 78. Lee Changkyu, ‘The Theory of Korean Finance’, 1996, p.l 14-132 79. Sangup, Hanil, Chaeil, Seoul Shintak, and Choheung were privatized. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 64 into the financial system, the government promoted the secondary8 0 financial institutions. Because of this privatization and promoting secondary financial institutions, firms were able to borrow more loans from these non-banking institutions and the Chaebols were able to partly own these non-banking institutions which were less controlled by the government compared to banking institutions. Secondly, in order to step up the banks privatization, the government abolished the 'Banking Institution Temporary Act' and revised8 1 the 'Bank Act7 as of January 1s t in 1983. It was to support the financial liberalization in terms of law. By decreasing the entry barriers, the competitions were encouraged. With foreign joint-ventures, two banks8 2 were newly set up. 12 short term financing companies and 58 mutual savings & finance companies, and 1 investment & trust companies were newly built up. Third, the government allowed the banks to engage in credit cards, factoring business, RP (Repurchases Agreement), CD, free savings deposit, and money trust business. It also allowed the non-banks to deal in factoring, RP, CP, CMA, and BMF. 80. Short-term financing companies (later it became merchant banks), insurance companies, Mutual Savings & Finance companies, investment and trust companies, securities companies, credit guarantee companies, leasing companies and etc except the banking institutions. 81. It included the abolition of the comprehensive order rights and the election rights of the bank executives of the Bank Inspection Board’s President. Second, the government set the voting rights stocks as 8 percent for the one legal person so that the large stock holders were not able to exercise the influence on the banks’ management. 82. Hanmi Bank and Shinhan Bank R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 65 Fourth, the government abolished8 3 the preferential interest rates for the policy loan in 1982. Interest rate band for the loan was introduced and 'Call Interest Rate" and 'CP interest rate' were liberalized8 4 in 1981. On January 12 in 1980, the government revised the 'Interest Rate Ceiling Act7 to increase the interest rate up to the market interest rate of maximum 40%. However, the financial liberalizations - the essentials are banks' self management and interest rate liberalization - were partly and namely carried out, not comprehensively. For the banks' self-management, the government continued to keep the indirect control of the banks. Monetary Inspection Board still had permission rights in naming bank directors, regulation rights, control rights, and punishment rights. The government was able to exercise the tremendous influence on the financial system through the Monetary Policy Committee and Monetary Inspection Board. Even though pre-approval procedures by the government for the bank budget, the settlements of accounts, wages, and office organizations of the banks were removed officially, the government continuously and customarily8 5 influenced the personnel management, settling the accounts and the wages of the banks. Until the mid-1980s, limited and partial interest rate liberalization was carried out, but strictly 83. This is an important step for the export support policy. 84. But, in 1982, the regulation for the CP interest rate was enforced again. The government reverted back to the regulation just after one year. 85. The government had the so-called ‘guidelines’ which played the critical role in the banks’ management decisions. The government influenced the decision of the bank’s president-naming, too. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 6 speaking, it was just in the nominal policy change. Actually the government continued to control the interest rate until the December in 19888 6 . Conclusively, during the Fifth Republic, the government still guided and controlled the financial market. Therefore, we would say it was the government predominant partly with market ruling model. (d) Export Support Financial System: lowering the export incentive parameter (1) The analysis of the ESFP changes: 1980-87 In the early 1980s, the international economic environment w asn't favorable to Korea due to the 1979 oil shock, high interest rates, higher raw material prices and new protectionism. 1970s heavy industry support policy caused higher inflation rates, economic concentrations, and ill-balances between the large enterprises and small enterprises. The new military government was trying to readjust the high growth focused heavy industry policy into the stable growth focused heavy industry policy. Therefore, it was necessary to cut down some of the favorable export supports of 1970s in the 1980s. The average annual growth was 7.4%8 7 which was lower than the 9.2 in the 1970s. Heavy industry share was larger than the light industry and the growth rate of exports surpassed the growth rate of imports in 1980s. Trade balances started to show the black letter from 1986 to 1989 for four years. 86. On December in 1988, except the loans with the government funds, all of the interest rates for the loans with bank funds was liberalized. We would say the real interest rate liberalization initiated in 1988. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 67 <Table2-19> Economic Data Contents 1980 1982 1984 1987 Average annual growth rate, % 1. The growth rate of the economy, % 2. Industrial Structure, % -2.1 7.2 8.2 11.0 7.4 - Agriculture/fishery 14.7 14.4 12.5 10.1 - manufacturing 3. Manufacturing structure 28.2 28.1 29.9 31.4 - Light 46.4 44.9 41.9 40.1 - Heavy 53.6 55.1 58.1 59.9 3. Export (million dollars) 17,214 20,879 26,335 46,244 9.2 4. Import (million dollars) 21,598 23,474 27,371 38,585 4.4 5. Trade Balances -5,321 -2,650 -1,373 9,854 Sources: Bank of Korea, Statistical Year Books, each year. On November 1 in 1985, three stipulations - export finance stipulation, foreign currency supply stipulation, and agriculture/marine product export preparation finance stipulation - in the export support financing was consolidated into one simple stipulation "Trade Loan Stipulation". In order to promote the exports of small and medium enterprises, the government introduced the 'Comprehensive Financing System', either. Since 1986, the Balance of Payments turned trade surplus and this trade surplus made the Korean economy inevitably open the goods and service market. Therefore, export support financing as a part of the policy loan had to be curtailed due to these trade surpluses and international pressures. From 1986, the won per dollar for 87. Percent change in GDP R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 8 export financing was decreased and on August in 1987, the government removed the finished product purchasing loan by trading companies. Other changes in each year were in table 2-20. <Table 2-20> Changes of Export Support Financial Policy Year Major Policy Note 1982 Abolishing the interest rate gap between the ESF loans and general interest rates Adjustment of ESFP loan rates 1985 Consolidating three stipulations of ESPF into the Trade Loan Rules: there are three name changes. • Export Loans — »General Export/Import Loans • Foreign Currency Supply Financing — > Construction Service Export Financing • Agriculture/Marine product Preparation Loans Procedure Rules — »Agriculture/Marine product Preparation Loans ESFP Loans Stipulations — * ■ Trade Loan Stipulations 1985 Introducing the Comprehensive Financing Systems in order to support SME exporters' exporting businesses In 1985, ESF Policies were changed in many aspects. 1986 Won per dollar for ESFP loans were decreased • Differentiating the unit price between the Large firms and SME firms: • Large firms: 670 Won per dollar • SME: 700 Won per dollar • Trade Surplus • Unit price decreased 1987 Won per dollar for ESFP loans • Large firms: 375 won per dollar • SME: 450 won per dollar • Abolishing the finished product purchase loans8 8 of Trading Companies Unit price decreased. Sources: Department of Finance, Bank of Korea, 1996 Because it was not possible for the government to provide the low interest rate export loan since 1982, in order to increase the exports, the government slowly increased the won per dollar for the loan and the loan volume until 1986. But the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 69 share of the export support loans in the total loans with banking funds gradually decreased. The share was 15.2% in 1980 and 5.8% in 1987. <Table 2-21> Status of ESF Loans (100 millions Won, %) ESF: A Total Loans with Banking funds: B (A/B)*100 1980 17,208 112,951 15.2 1981 21,972 147,633 14.9 1982 22,784 187,183 12.2 1983 26,200 222,107 11.8 1984 27,654 262,678 10.5 1985 31,299 313,108 10.0 1986 34,445 366,364 9.4 1987 24,204 418,734 5.8 Sources: Bank of Korea, Monthly Economic Bulletin Comments: 1. status of ESF Loans is the Trade Loans8 9 in the deposit money banks 2. B: Loans with banking funds. The loan price per dollar was 530 won per dollar in 1980 and since 1986, the government applied the different unit price per dollar for the large firms and small firms. Table 2-21 shows the export support loans and the share in the total loans in each year. Table 2-22 shows the won per dollar for the export loans in each year. <Table 2-22> Won per dollar for Export Financing Loans Unit: Korean Currency Won Year Unit price per dollar for the loan Contents 1980 530 1982 570 1983 655 1985 740 SME 740 Non-conglomerated Large firms 740 Other Large Enterprises 88. This loan was to promote the benefits of the domestic export goods producers. 89. Trade Loans include general export/import loans, construction service loans, and agriculture/fishery loans over this period. It does not include the export-equipment support loans. Bank of Korea, Monthly bulletins. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 70 1986 700 SME 670 Non-conglomerated Large firms 670 Other Large Enterprises 1987 520 SME 375 Non-conglomerated Large firms 175 Other Large Enterprises Source: Bank of Korea, Monthly Economic Bulletin, each month. Table 2-22 continued. There were important changes in the interest rates for the export support financing loans. Since 1982, the interest gap between the ESF Loan and General Loan rates was removed by the government. The government was not able to provide the lower interest rates for the exports due to the international pressures. The spread between the ESF Loan rate and general loan rate was 7.7% in 1980, and 4.5% in 1981. The interest rates of the export support financing were 11.8% in 1980, 15.0% in 1981, and 10.0% in 1986. The average ESF Loan rate was about 10.9% over the period. <Table 2-23> the spread of interest rates between ESF and General .oans, % ESF interest rate General loans rate Spread: B-A 1980 11.8 19.5 7.7 1981 12.0 16.5 4.5 1982 10.0 10.0 0.0 1983 10.0 10.0 0.0 1984 10.0 10.0-11.5 0.0-1.5 1985 10.0 10.0-12.5 0.0-2.5 1986 10.0 10.0-12.5 0.0-2.5 Sources: Bank of Korea, Economic Statistic Yearbook, each year, p.64-65 Comments: 1. ESF Loans: the interest rates on loans for Foreign Trade Bills of Deposit M oney banks in the commercial bank. 2. General loan rates: loans on installment savings (1) The Characteristics of the ESFP Loans in the third period: 1980-87 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 71 The government still guided the economic development plans and controlled the financial markets strongly. However, this period turned adjustment period of Export Support Financial Policy. South Korean government was no longer immune to the international pressures. There were a lot of international pressures from GATT, U.S. 1979 Trade Agreement Act, U.S. 1984 Trade Tariff Act, World Bank and U.S. regulation of Super 301 against South Korean economy. Therefore, in this period, the government had to respond to the international pressures and change the Export Support Financial Policy. The government could not provide the lower interest rates for export support and decreased the amount of ESFP loans in the middle of 1980s. For example, the government decreased the won per dollar for export support loans to 375 won in 1987, and repealed the finished product purchasing loans by Trading Companies. The interest rate for the export loans was about 11% which was lower than the general loan rate 12% in the average over the period. Compared to the 1970s, one of the important benefits was no longer available to the firms. The preferential interest rates were not there since 1982. <Table 2-24> the Features of ESFP Loans in 1980-1987 Contents Features Supporting system Abolishing the purchase-financing system of finished products of Trading Companies Decreasing the won per dollar for export support loans Interest rates Since 1982, the interest rates gap becomes zero R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 72 The amounts of ESFP Loans Average total loan in banking funds was about 25 trillion won and average ESPF loan was 2572.1 billion won. - Average share of ESFP to the total banking funds was 11% compared to 14% in the second period and 8.5% in the first period.____________________ Table 2-24 continued. 4. Structural Adjustment & Openness and Globalization 1988-1997: Price reform Period (a) International Political Economy : high pressures to liberalize Around the times of the 6th Republic9 0 , international economic system has been changing rapidly. The major changes came from these sources: (1) the influence from WTO and OECD membership, (2) New UR round, (3) Regionalism in Europe, Asia and the Americas. In 1980s, as we studied in previous sections, the world experienced lower economic growths, higher unemployment rates, and trade imbalances9 1 between the advanced countries. Each advanced country had a tendency to have the government-managed trade9 2 instead of free trade to cope with those economic problems. Among the advanced countries, there came a new multi-lateral round so-called "UR round" to solve those problems of the government-managed trade system in 1986. This UR round was the most comprehensive and multi-lateral agreement and 9 0 . President Roh Taewoo government which started On Feb 25 in 1988. 91. Japan and German enjoyed the huge trade surplus and U.S. suffered the huge twin deficits in 1980s. 92. It means that more positive policies for the export and negative policies for the import by the government. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 73 it dealt with: (1) lowering the tariffs and removing or decreasing of the non-tariff barriers, (2) strengthening trade dispute- solving procedures and the functions of the GATT, (3) setting up new trade agreements which included intellectual property, services, and trade related investment measures. On Dec 15 in 1993, new international trade system was imposed and on January 1st in 1995, World Trade Organization (WTO)9 3 was created. Due to the new countervailing measures and subsidy stipulations of WTO, Bank of Korea's re-discounts on commercial bills, the subsidy based on the export results and different won/dollar loan unit for the imported raw materials for the exports had higher possibility of being classified as the subsidy. In 1996, Korea won the membership of the OECD, and related to this membership, the Korean government had to have the capital market liberalization and financial market opening. Capital market liberalization9 4 gradually allowed one 93. The characteristics of WTO : (1) New Trade Areas: Agricultural Products, Services, Intellectual Property Rights, Trade related Investment, (2) Regulated the new stipulations to overcome the limitation of the GATT. South Korea won the membership of WTO in 1995, too._______________ Contents: differences GATT WTO The character International Agreement Legal personnel: permanent department (DSB: Dispute Settlement Body) which solves the trade conflicts Tariff and non-tariff Only lowering the tariffs. Lowering tariffs and abolishing the non tariff barriers Trade areas Most o f them are manufacturing goods Agricultural products included New Regulations No stipulations o f services, intellectual property rights Services, intellectual property rights, investment measures service agreement contracted international norm was made for intellectual property rights: patent rights, copy rights, trade mark rights Sources: Korea Trade Association, “Trade Yearbook”, 1995 94. Stock and Bond market. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 74 foreigner to purchase the stocks of a company up to 10% till December 2000. If the conditions9 5 allow, bond market will be fully liberalized for the foreigners. For the financial market opening, favorable M&A became possible by abolishing the ownership limit of the foreigners to the non-banking institutions. The new business9 6 areas of the foreign banks were allowed. Life insurance market was open to the foreign companies with the joint-ventures since 1988. Another international economic change came from the blocked regionalism such as EU, NAFTA, and APEC. Within the regional membership, the membership countries pursed the free trade and increased trade barriers against outside membership. For the case of APEC, 18 APEC countries declared the 1994 Bogor Summit Conference, within the region, the advanced countries would finish free trade and investment until 2010 and the developing countries would finish them until 2020. U.S. with the 1974 Trade Act, 1988 General Trade Act, and bilateral agreement, cancelled the GSP(Generalized System of Preference) for South Korea in 1989 January. U.S. criticized the South Korea as the exchange rate manipulating country in 1988 and asked and pressured the government to adopt the free market exchange rate system in 1990. With the super 301 section in the Trade Act, in 1988, U.S. 95. Macroeconomic stabilizations are met or if the interest rate gap between the domestic and international rates is below 2%. Sources: Kang Moonsoo, ‘OECD membership and Policy Response of the financial sector’, 1997, KDI, p.5 96. (1) Increasing the number of branch, (ex) City Bank: three more branches in 1988. (2) Bank of Korea’s rediscounts and loans - trade loans (3) CP businesses. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 75 pressured the Korean government to open the tobacco, beef, and grape markets. Based on 1988 General Trade Act, U.S. acquired the protection of the intellectual property rights and trade mark rights from South Korean government in 1990. There were continuous pressures from U.S. government for South Korea to liberalize the market. Since 1988, the pressures to open goods and financial markets had been in increase strongly and gradually. The Korean government had no choice but to follow the demand from the international economy and to arrange the liberalizations and openings. (b) A Relationships of Government and Firms Since the 6th Republic9 7 , firms did not just follow the government's economic policy any more. They expressed their interests directly to the government and sometimes were against the government policy which might have negatively affected the firms' interests. For example, when the 6th Republic intended to control the large firms with "Land Public Concept Act", "5.8 Non-business Real Estates Forced Sale Measure", and "Regulation for the Ownership Concentration", the chaebols did their best to weaken the original ideas for those laws. Some chaebols even brought the civil lawsuit9 8 of constitutional rights against the government. 97. Since 1987 June Democratic Movement, Korean people achieved the direct presidential election system. December in 1987, people voted for the President directly, and the 6* Republic initiated new democracy oriented period. 98. Kang Sukyoo 1 , “the relationship change between the state and the chaebol,” Yonsei University, 1994, p71. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 76 Even the president of the Hyundai Group 'Jung Juyoung' created the 'National Party (Kukmin dang)'and participated in the Presidential Election in 1992. His party earned 31 seats in the national assembly in 1992 March General Election. The government even did the internal revenue service investigation to prevent the political money from coming into his party, but it did not find any other cause to punish the chaebol's participation in politics. The following figure <figure > showed how the firms did not follow and even did bring the lawsuit against the government. Due to the rapid development of the stock market since the mid-1980s, with the democratic social environment, the firms financed substantial portion" of the loans from financial market, not through the financial intermediaries. <Figure 2-1 > Influencing Relationships between the government, banks and the firms 1960, 70s and early 1980s: Government controlled and guided the banks and the firms Firms borrowed the loan from banks & followed the Banks are under the govemme leadership 99. Firms’ fund-raising structure change Unit: % Contents 1980 1985 1988 1990 1993 Indirect financing: bank or non-banks 34.5 48.5 24.2 38.4 31.4 Direct financing: stocks, CP, bonds 22.0 26.2 52.6 42.4 49.1 Others: foreign debt + the rest o f others 43.5 25.3 23.2 19.2 19.5 Source: Bank of Korea, ‘Understanding of the Money Flow in Korea’, 1994, p.73 R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 77 / Since mid-1980s: \ / Stocks and \ / Firms financed the \ / bonds market \ loanable funds \ were liberalized ' through stocks and I with foreigners bonds market: •4— — ► participating listened to markets / not the government Figure 2-1 continued. Therefore, the government had no choice but to allow the market-ruled mechanism and the relationship between the firms and the government were not just order & follow one. Government still guided in some senses, but sometimes the government and the firms expressed different interests and concerns and conflicted with each other. There are objective functions of heterogeneous interests from two different economic decision-makers. Because of 1986 trade surplus and 1989 expansionary economic policy, the quantity of money stock greatly increased in short period and this made the monetary policy harder to control the money stock effectively. Interesting feature related to this money stock increase was firms' management of flow of funds. Despite the fact that financial deficits to GNP decreased1 0 0 and the ratio of the 100. Firms’ financial transactions and usage, Unit: % 1985 1986 1987 1988 Financial Deficits/GNP 10.1 7.6 6.8 5.7 Equity Capital/Equipment and Inventory Investment 75.7 70.2 78.9 102.9 Land Purchase/funds Employed 3.3 2.9 3.9 5.7 Source: Bank of Korea, Financial Transactions of the Firms, each year. Lee Changkyu, Theory of Money, p. 126 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 78 internal funding to the equipment and inventory investment increased, once the government support for the export support loan was changed negatively to the firms and international economic environment was changed in that same way, the firms' outside funding1 0 1 amounts started to increase with the increase in the land purchase. This probably means that the firms used their extra money or outside funding to purchase the high return financial assets or real estates. <Table 2-25 > Finns' Net Purchase of the land and Total Loans, Unit: 1 billion Won 1985 1987 1989 1991 1993 1995 Net purchase of the land 518.9 1084 2405.4 4459.8 5747.7 10948 Total loans 7120.4 7202.8 13660.7 24343.1 20373.2 31854. 8 Sources: Bank of Korea, Financial Transactions, each year. The above table 2-25 shows that amount of the land net purchase and the total loans for the firms. It seemed to have a strong correlation1 0 2 and significant relationships between the firms' land purchase and the total loans. This figure might suggest that moral hazard firm may choose higher level of effort in undesirable business such as purchase lands and may choose lower level of effort in exporting goods and service in the competitive international markets. (c) A Relationships of Government and Financial System Due to the financial market liberalization policy since the mid-1980s, the financial institutions have been growing rapidly and financial market liberalization has been 101. Bank Loans + stocks + bonds. Outside Funding increase will increase the firms’ debts (except the stocks) 102. The correlation coefficient between these two variables is more than 0.87. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 79 carried out. Since the 6th Republic, the financial institution has been moving the center from the government to the financial markets - capital markets such as stocks, CP, and corporate bonds markets. More banks were newly set up, and the secondary financial system was grown rapidly which resulted in the change of the firms' fund-borrowing pattern. Bank of Korea's independence from the government has been enhanced. The minimum self-management of the banks to name the president of the banks was acquired. The interest rate liberalization was carried out fully in 1991. The relationship between the government and the financial system changed into the capital market initiative system, not the government initiative system. The government relaxed the restrictions of setting up the new banks since 1988. Dongwha Bank, Dongnam Bank and Daedong Bank were newly opened. For the secondary financial system, five local investment & trust companies, 11 local leasing companies, 24 life insurance companies and about 40 non-banking institutions were newly opened up. Since 1986, due to the trade surplus, the government revised the 'Capital Market Growth Act7 , promoted the capital market with asking more firms to open and encouraged the listed firms to raise the funds through the issue of new shares to be purchased. Since 1986, because the trade surplus continued to exist, the price decontrol of the interest rate became the important issue. In 1988 December, the government R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 8 0 comprehensively initiated the interest rate liberalization for the loans except1 0 3 the loans with government funds. For the deposit rates, only the long term rates were liberalized. Even though the government reverted back to the regulation of the loan rates after one year later due to the pressures to support the expansionary policy, it could not avoid the interest rate liberalization as the economy became more open. Therefore, in 1991 August, the government announced the four-step interest rate liberalization plan1 0 4 and carried out the plan on the schedule. Therefore, the government-financial market relationship was on the move to the capital market initiative system. (d) Changes in Export Support Financial System (1) The Analysis of the ESFP changes: 1988-97 In the late 1980s, the Korean economy experienced high economic growth rates, trade surplus and lower inflation rates. To increase the industrial competitiveness, the structural adjustment was continuously brought in and the R&D and the SME (Small & Medium Enterprises) received a lot of policy attentions. The growth rate of the economy was over 8%, and the growth rate export was two digits. But since 103. Exception: funds with the government, NIF, and Bank of Korea’s agriculture/fishery/livestock loans. 1 04 . <Interest rate liberalization plan>________________________________________________________ Stages Plans 1st: 1991-92 Banks Overdrafts, commercial bills discounted, CP and Trade loan discounted 2n d : 1992-93 Banking and non-banking: all loans 3r d : 1994-96 The funds with government and BOK’s rediscounted, some short-term deposit rates 4“: 1997- All the deposit rates, all government bonds R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 81 1990, the economy went back to the trade deficits, again. Table 2-26 shows the general economic data to show the trend. <Table 2-26> Economic Data Contents 1988 1990 1992 1994 Average annual growth rate, % 1. The growth rate of the economy, % 2. Industrial Structure, % 12.0 9.5 5.1 8.6 8.7 - Agriculture/fishery 10.2 8.7 7.4 7.0 - manufacturing 3. Manufacturing structure 32.1 29.2 27.8 26.8 - Light 36.8 34.1 30.6 26.9 - Heavy 63.2 65.9 69.4 73.1 3. Export (million dollars) 59,648 63,12 4 75,169 93,676 15.3 4. Import (million dollars) 48,203 65,12 77,316 96,822 16.1 5. Trade Balances 14,161 7 -2,179 -4,529 -3,146 Sources: Bank of Korea, Statistical Year Books, each year. For the export support financing policy, in 1988, the government abolished the export support loans to the Large Group Enterprises, and even decreased won per dollar loan price to the SME and non-grouped large enterprises. 30 chaebol groups were not able to use the export support loans which were provided by the banks. This was one of the serious changes. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 8 2 However, in 1990 and 1991, to solve the trade deficit problems, the Korean government took the efforts of more export support loans for the SME, non-grouped large enterprises, and commercial banks trade bills discounted1 0 5 to be allowed. This effort faced the opposition of the foreigners and caused problems of rapid growing money stocks. <Table 2-27> Changes of Export Support Financial Policy Year Major Policy Note 1988 Abolishing the trade loans to the large enterprises 1994 BOK adopted the Total Limit Loan System and absorbed the Trade Loan System since World Trade Organization systems started in 1994. Trade Loan is treated like a general loan. Direct Support — * ■ Indirect Support 1995 Full Liberalization of the interest rates on ESFP loans Since 1992, due to the effect of Interest Rates Liberalization, ESPF Loan was a bank prime rate, and in 1995, it is fully liberalized. Sources: Department of Finance, Bank of Korea, 1996. As WTO set off, in 1994, because there were concerns that the export support financing loans were considered as the subsidy under this WTO system, the ESFP loan was changed from the automatic financing support by the export cases into the Total Limit Loan System1 0 6 . The ESFP loan system was treated like a general loan since then. 105. 30 chaebols use this loan system to finance the loan for the exports. The feature of this trade bills discounted: (1) non-policy loan and it is provided by the private sector, (2) interest rate is determined by the market, (3) with L/C or local L/C, the firms borrow the loan easily. Source: Bank of Korea, 2003 106. Since March 15th in 1994, this has been one of the loan systems to the banks by the Bank of Korea. Based on the commercial bills discounted, and trade loans of each banks’ loan amounts, the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 83 The size of the ESFP loan decreased between 1986 and 1988 due to the trade surplus, after 1990, it was gradually increasing because of the trade deficit. But the share of the ESFP loan in the total loan with bank funs has been decreased from 2.6% in 1988 to 1.6% in 1996. <Table 2-28> Status of ESF Loans (100 millions Won, %) Year ESF: A Total Loans with Banking funds: B (A/B)*100 1988 12,016 457,132 2.6 1989 13,822 547,777 2.5 1990 19,473 683,791 2.8 1991 22,543 806,325 2.8 1992 25,422 957,464 2.7 1993 24,734 1,093,367 2.3 1994 27,113 1,269,055 2.1 1995 28,469 1,448,786 1.9 1996 26,793 1,654,173 1.6 Sources: Bank of Korea, Monthly Economic Bulletin Comments: 1. status of ESF Loans is the Trade Loans1 0 7 in the deposit money banks 2. B: Loans with banking funds. Since 1988, the ESFP loan was provided to the SME and non-grouped large enterprises. Loan price per dollar was 450 won for SME and was 200 won for non grouped large enterprises in 1988. Loan price has been increase gradually up to 700 Bank of Korea provided the loans to each banks within the limit of the total loans which is determined by the Monetary Committee. (1) Purpose of the Total Limit Loan System: let the Bank of Korea pursue its goal of controlling the quantity of money and promote the loan support for the SME, export firms, and local firms. (2) Because the interest rate (BOK charges 3% for the banks) was lower, the banks use this loans for the SME support (3) The support methods are by the banks and by the BOK branches. Source: Science & Industry Information Service, 2003, ‘Financial Information’ 107. Trade Loans include general export/import loans, construction service loans, and agriculture/fishery loans over this period. It does not include the export-equipment support loans. Bank of Korea, Monthly bulletins. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 84 won for SME in 1996. Also the loan price1 0 8 was applied a little differently by the funds' usage such as export goods production loan, raw material financing, and comprehensive1 0 9 financing. <Table 2-29> Won per dollar for Export Financing Loans Unit: Korean Currency Won Year Unit price per dollar for the loan Contents 1988 450 SME 200 Non-affiliated Large Enterprises 1989 550 SME 300 Non-affiliated Large Enterprises 1990 600 SME 400 Non-affiliated Large Enterprises 1991 650 SME 400 Non-affiliated Large Enterprises 1992 650 SME 400 Non-affiliated Large Enterprises 1993 700 SME 400 Non-affiliated Large Enterprises 1996 700 SME 400 Non-affiliated Large Enterprises Source: Bank of Korea, Monthly Economic Bulletin, each month. Table 2-29 continued. In the table 2-30, the loan rates of the export support financing was 10.0% in 1988, 8.5% in 1993, and 10.5% in 1996. The average ESF Loan rate was about 9.8% over the period. The gap between the ESF loans and the general loan rates got smaller and since 1992, due to the Interest Rate Liberalization, the ESF Loan rate was 108. In 1997 February 6th , it is introduced that the New Calculation Method for the Loan Price per dollar for export loans by the BOK. Source: BOK, Finance Department, 1997. 1 0 9 . Comprehensive Financing: based on the previous one year export amounts, if the export was less than 50 million dollars, regardless of the fund’s usage, all of the needed loans were provided. This is designed to promote the exports of especially SME under the Total Limit Loan System. Source: Bank of Korea, 2003, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 85 determined at the bank prime rate. Even in 1995, the ESF Loan was fully liberalized1 1 0 . <Table 2-30> the spread of interest rates between ESF and General Loans, % Year ESF interest rate General loans rate Spread: B-A 1988 10.0 11.0-13.0 1.0-3.0 1989 10.0 10.0-12.5 0.0-2.5 1990 10.0 10.0-12.5 0.0-2.5 1991 10.0 10.0-12.5 0.0-2.5 1992 10.0 10.0-12.5 0.0-2.5 1993 8.5 8.5-10.0 0.0-1.5 1994 9.0-9.5 9.5-12.5 -0.5-3.0 1995 9.5-10.5 9.0-12.5 -0.5-3.0 1996 10.5-11.0 11.0 0.5-0.0 Sources: Bank of Korea, Economic Statistic Yearbook, each year, p.64-65 Comments: 1. ESF Loans: the interest rates on loans for Foreign Trade Bills of Deposit Money banks in the commercial bank. 2. General loan rates: loans on installment savings (2) The Characteristics of the ESFP Loans in the third period: 1988- 97 Because of the international pressures such as OECD memberships, new multinational UR round, WTO, and international banking institutions, the government had to accept the pressures from the foreign to open the economy. Therefore, at the table 2-31, in 1988 the government abolished the export support financial loans to the 30 large grouped enterprises, and in 1995, liberalized the interest rates on the ESFP loans. ESFP Loans system (Trade Loans Rules) was absorbed1 1 1 into the Bank of Korea's Total Limit Loan System, too.1 1 2 It is not 1 1 0 . in November 1s t 1993, each bank was able to determine the interest rates. Since 1995, the bank determined the ESFP loan rates with the bank prime rate + added rate (0.5%~1.5%) 111. Direct Support System — > Indirect Support System R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 8 6 automatically provided any more by the export cases. The loan amount to the each bank should be within the total loan limit. Therefore, based on the Total Limit Loans determined by the Monetary Committee, the ESFP loan was provided to each banks based on its loan results by the Bank of Korea in advance. The banks which borrowed the ESFP loan from the BOK determined the interest rates for the ESFP with the added rates (0.5-1.5%) in addition to the bank prime rate1 1 3 . Comprehensive Loan system was introduced in order to promote the ESFP loan for the SME which used to have the smaller amount of the exports. The ESFP loan system became favorable to the SME and non-grouped large enterprises. For the 30 chaebols, they were able to use the trade bills discounted easily. These trade bills discounted are not subject to the government policy loan, but are provided by the private financial institutions such as private banks, and merchant banks. The interest rate for this trade bills is determined by the market. The share of the ESFP loan in the total loan was on the decrease gradually, but the size of the ESFP loan was on the increase most of the time during this period. <Table 2-32> The Features of ESFP Loans in 1988-97 Contents Features Supporting system abolishing the ESFP loans to large enterprises Trade loans was absorbed in the BOK Total Limit Loan System Indirect export support system Interest rates There was no gap between the Export Loan rates 112. ESFP loans were not subject to the total limit of the loans before. It is incorporated into the total limit loan system since 1994. 113. The bank prime rate was 8.5% in 1997 and as of January 1s t in 1998, it became 11.5%. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 87 and general loan rates. The amounts of ESFP Loans Average total bank loans were 99,087 billion won and ESFP Loans were 2,226.5 billion won. The share of ESFP was just about 2% over the period Table 2-32 continued. D. Comparing pre- and post-structural adjustments centered on four different time periods. The following tables are the comparisons of all periods for the changes of the Export Support Loans and the related issues to this ESFP loans. 1. Government, International pressure and Government-Firm relationship <Table 2-33> Pressures from International Institutions, and relationships with firms International pressure to openness Government-Firm relationship 1962-1971 Very low High over firm 1972-1979 Low Very high over firm 1980-1987 High High over firm Since 1988 Very high Low over firm 2. Changes in ESFP in different four periods <Table 2-34> Comparisons of Export Support Loan Systems ESFP Strength Changes of the characteristics of ESFP 1962-1971: Export Orientation policy Strong policies were introduced Introducing various ESFP measures 1972-1979: Export Priority policy Very strong policies were introduced and extended Integrating & extending ESFP system: Effectively Support 1980-1987: adjustment of Export Priority Policy Policies were weakened ESFP waning: repealing the purchasing loans by Trading Company, and the won per dollar for the ESFP Loans decreased. Since 1988: Openness, Very weak policy No longer ESFP system: Abolishing the ESFP loans R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 8 8 Segaehwa to large firms. Direct ESFP (Globalization) measures disappeared. Trade loan system is absorbed into total loan limit system of bank of Korea. Table 2-34 continued. <Table 2-35> Comparisons of Interest rates and amount of loans Unit:100 million Won, % Periods Average Interest Rates, % The volume of the Loans Export General Spread = General - Export Total Bank Loans Export Loans Share of export loans, % 1962-71 7 22 15 3,360 129 5 1972-79 8 17 9 29,605 5,241 14 1980-87 11 12 1 253,845 25,721 11 1988-96 10 10 0 998,740 22,265 2 <Table 2-36> The Analysis of the International Pressures, government leadership and policy tools._______________________________________________________ Periods International Pressures Government Leadership in export support policy Policy Tools 1962- 1971 -light industry - export- orientation Very low Strong government leadership Strong 1972- 1979 - heavy industry - export priority Low Very strong governm en t leadership Very strong 1980- 1987 adjustment period High er Responsive to the international pressure: weak Weak R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 89 1988- Openness/ Very Responsive to the Very weak 1996 globalizati high domestic and on international pressures: very weak Table 2-36 continued III. Empirical Analyses A. Conceptual Framework The purpose of this paper is not to estimate the parameters in the structural behavior equations of the firms' decisions and government's incentive functions, but to present the inter-relationships between the economic variables such as the exports, export support loans, the export support loan interest rates and the other influencing variables in the frame of international political economy approach. Within heterogeneity of interests and moral hazard model, we will investigate how the government financial reform policy has been interacting with the firms' decisions to exports. Therefore, among the changes of inter-relationships, we will see how incentive changes given by the government's Export Support Loan Policy increase or decrease firms' commitment of exports depending on the periods of pre- and post-structural adjustment. We will also see how international political economy factors affect the government's policy commitment so that this influences the firm's commitment on exports. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 90 The objectives of this chapter are two folds: (1) how the government's financial reform is interacting with the firms' commitment in the political economy structure. (2) so that we will determine the relative importance of incentive changes given by industrial economic policy variables of the endogenous variables in the time series econometric models. We'll directly test the structural changes in the government policy change and export commitment change under the different international political economy factors. Then we use a standard VAR and structural VAR models to determine the influences of different incentive system on export and ESFP loan commitments. Therefore, we can prove the model's argument that firms react to less Investigating the relationships between the variables would result in helping understand the structural relationships of those variables with the political economy approach. The structures and importance of this empirical chapter is the followings; (1) data will be explained. (2) With a dummy variable analysis, whether there is a structure change will be analyzed. We expect the dummy variable of structural adjustment period will be significant so that we confirm there is a change of incentives on exports commitment and ESFP loans. (3) Based on the result of Hausman specification test, standard VAR model will be introduced. This reduced form VAR has benefits of treating all variables endogenously so that we will see how each economic variable influences each other's commitments of exports and ESFP loans. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 91 A standard VAR of Pre- and post-structural adjustment analysis make us compare the size of proportions of FEVD (Forecast Error Variance Decomposition) for the periods of pre- and post-adjustment. We consider the size of the proportions of each variable to the innovations of other variable as the proxy for importance of each variable. If we find smaller proportion of ESFP loans on exports in FEVD after the structural adjustment, w e'd say exports is less motivated by the government incentive system of ESFP loan after the structural adjustment. Because this standard VAR has Cholesky Decomposition of triangular restrictions on the structural parameters, it does not reflect any structural theoretical assumptions of this original model. Even though the results of a standard VAR help us compare the pre- and post-structural relationships of the variables, it is really worth while to introduce the Structural VAR (SVAR) model since SVAR reflect the model's assumptions in the estimation procedures. Introducing VAR and SVAR gives us better understanding of the incentive and commitment changes of all variables over the time span because they deal with all variables endogenously and SVAR reflects the model's assumptions for the structural relationships among endogenous variables. Those will enhance the model's applicability and convincing power. B. Data R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 92 Most of the data used in this analysis are from the Statistical Year Book and Monthly Bulletins published by the Bank of Korea, U.S. Bureau of Economic Analysis, Korea National Statistical Office and the International Financial Statistics Book. The data sets are composed of annual and quarterly data sets on the South Korea's exports and export support loans interest rates. The sample period is the annual 1962-1996 and the quarterly period is from 1967.1 to 1997.4. EPI indexes are available from 1st quarter in 1971. Openness is calculated by the exports and imports relative to the GDP. The variable "Press" is the number of measures what the U.S. government took as import regulations of Korea's goods. These import regulations are composed of anti-dumping1 1 4 controls, countervailing duties, safeguard1 1 5 , and the bilateral agreements of two governments. Anti dumping has been taking more than 60 % out of all U.S. import regulations to Korea. In the econometric Analysis, one of the issues is whether the researcher has to use the nominal data or real data. For the Export Support Loan procedure, there are two forms <Data table > History of the data ________________________________________ Variables Note Nominal Exports U.S. Current Dollars, Korea National Statistical Office Real Exports U.S. Constant Dollars in 1996 = 100 1 1 4 . Most of the U.S. import regulations have been in the anti-dumping stipulations to Korea. This has been in increase since the mid-1980s. KIEP, 1997, “The Political Economy of Korea-U.S. Trade Conflict”, pp.l -10. 1 1 5 . U.S. import regulations are composed of two broad areas; (1) Import Restrictions, (2) Unfair Trade Control. Safeguard is one of the import restrictions based on the Trade Act Article 201 that is to restraint the import itself. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 93 Nominal Export Support Loans Korean Currency Unit: Won Real Export Support Loans Korean Constant Won in 1996 = 100 Nominal Export Support Loan Interest Rates In the Commercial Banks Account, Bank of Korea Korean GDP deflator 1996 = 100, IFS Wage Index International Financial Statistics Data Export Price Index: EPI 1996 = 100 1. Bank of Korea, Dollar base price index 2. IMF, International Financial Statistics U.S. GDP deflator 1996 = 100, U.S. Bureau of Economic Analysis International Political Economy: Foreign Pressures, PRSS U.S. regulations for the imports from Korea: Includes anti-dumping measures, countervailing duty, and other measures and penalties from case studies and the chronology analysis in the newspaper Source: Bank of Korea, National Statistical Office, Ministry of Finance, U.S. department of Commerce, IMF. Data Table continued. of loan procedures. The one is the L/C based loan process and the other is previous export volume base process. When the banks lend the loan, the banks used the nominal base dollar-term export amount. Therefore, it is worth while to investigate the inter-relationships between nominal export policy loan1 1 6 and nominal exports as well as the relationships of real variables. All of the data we are using are time series data. Empirical work based on the times series data assumes that the underlying time series is stationary. We tested whether the variables are stationary or non-stationary. In related to this non- stationary, autocorrelation will be taken care if the underlying time series is nonstationary. Doing the regression analysis with the non-stationary time series will lead to the spurious or nonsense regression. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 94 I used the Dickey Fuller and Phillips Perron Test. Augmented Dickey Fuller test and Phillips-Perron test show similar results that the time series macro variables - ESFP loans, and the Exports - have unit roots1 1 7 . We take the first differenced data which become stationary for the analysis. C. Investigating the inter-relationships in Structural Change and Simultaneity W hat we investigate here is fifth different steps to see the inter-relationshps among variables. We will test, first, whether there is a structural change in the government's reward system. We will see the varying dummy parameter estimate to see the change of government reward system. Second, if there is a change of incentive parameter '(3', we will do the analysis of the variance for the growth rates of exports depending on the period before the structural break and after the structural break. From this ANOVA, we will see whether firm's commitment to exports is significantly different in these two different periods - before and after structural break. And we will test how export7 s commitment is influenced by new structurally-changed loan institutions. We expect that even after the structural break in the export support loan system, ESFP Loan should be still significant but the interaction dummy "DMGLO (Dummy * GLO)1 1 8 " influences negatively on the export7 s commitment. 116. Thus, the following empirical works are for the nominal ones which show the similar results as the real variables. 117. All of the Econometric Analysis is done by the SAS 8.0 version and the detailed unit root test is available upon request. 118. Dummy Variable * Growth rate of ESFP loan: Dummy for Structural break R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 95 Third, we will use the polynomial distributed lag model where the change in export has an effect on the dependent variable 'Export Support Financial Loans' that is distributed over several future periods and the change in export support loans has an effect on the dependent variable 'exports'. Fourth, we raised the issue of the Simultaneity which is embedded in the theoretical model. The firm and government make the decision as each other's decision given. This raises the endogeneity problem and it is tested by the Hausman specification test. Fifth, we introduced a Vector Auto regression (VAR) to look at the inter relationships between the decision variables by the government and firm. The Forecast Error Variance is presented to show how the firm's commitment to exports is influenced by the government policy and how the government policy loan was affected by the firm's different commitment over the periods of before-the- structural break and after-the-structural break. Introduction of the forecast error variance decomposition to deal with the changes of incentive of firm and government is a unique analytical development in this area of time series econometrics and development economics. 1. Structural break in the incentive parameter and Analysis of Variance (a) Structural Change in Incentive Parameter The government eliminated the gap between the ESFP Loan and general loan rate in 1982. Since 1988, the government was determined not to provide the ESFP R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 96 Loans to the large grouped enterprises1 1 9 . Only SMEs received the ESFP loans since then. In 1994, ESFP Loan system was absorbed into the Bank of Korea Total Limit Loan System1 2 0 . ESFP loan rate1 2 1 was fully liberalized in 1995 as a part of the interest rate liberalization policy. Therefore, since the mid-80s to the late 1990s, we would guess the structure of the economy is pretty different from the previous periods of 1960s and 70s. We created the dummy variable to check whether there is a structural change in the incentive parameter over the periods since 1986. We will see how these changes are reflected in the varying dummy parameter called "PDEX". The equation is, DLO = cc + PD EX + yPDEX (3-1) We tested the equation 3-1 whether there are structural changes in the government incentive system for the firms' commitment to the exports. We created the new interaction variable 'PDEX = DUMMY*DEX'. South Korea experienced trade surpluses in 1986,1987,1988, and 1989. International pressures to open the 119. Since Feb 1988, trade loans and export industry equipment loan became available only to the SMEs and nonaffiliated large enterprises. 120. absorption into the Total Limit Loan System suggests direct export support policy is changed into indirect support p olicy since then. 121. Even though the government removed the preferential rate gap since 1982, it kept maintaining the rate at the bank prime level by the indirect guide to the banks until 1994. In 1994, the ESFP loan was transformed into the indirect support system due to the Total Limit Loan System. In 1996, WTO (World Trade Organization) concluded the Korean ESFP was not a subsidy, but it still has in itself a cause of problem depending on the change in the ESFP Loan Policy. After 1995, by the window guide of the government, the rate was determined with the bank prime rate + additional margin. Therefore, until it was fully liberalized, ESFP loan interest rate was not fully determined by the market for the loan. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 97 economy were put into the policy makers throughout the 1980s, especially since the mid-1980s. <Table 3-l> Structural change in the incentive parameter Dependent variable: DLO, the first differenced loan, R-square = 0.2816 Variable Parameter estimates t- value p-value Intercept 889.93662 0.91 0.3700 DEX 0.60500 1.43 0.1619 PDEX -0.92773 -2.30 0.0281 1. test result is corrected for serial correlation in errors The result shows at table 3-1. We expected the significance of the DEX would be decreased and the negative estimate of "PDEX" became significant. In the above table, as expected, the parameter estimate of the DEX became insignificant and PDEX became significantly negative at 5 and 10 percent significance level. We checked if there is a structural break in the incentive parameter in the growth rate of ESFP loans. We created the varying parameter "DMGEX = DummyxGEX1 2 2 " and the significance test. The equation 3-2 is, GLO = a + y0D M + J3GEX + yx DMGEX (3-2) The result is as follows at the table 3-2. <Table 3-2> Structural change in the incentive parameter in the GLO Dependent variab e : GLO, the growth rates of ESFP loans, R-square = 0.4414 Variable Parameter estimates t- value p-value Intercept 0.1071 0.87 0.3893 DM 0.1160 0.68 0.5031 GEX 0.7932 2.06 0.0487 DMGEX -2.6114 -3.06 0.0047 122. GEX = Growth Rates of Exports, GLO = Growth Rates of ESFP Loans R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 98 1. This is corrected for the serial correlation. 2. DM: the constant dummy. As we expected the structural break in the incentive parameter, the coefficient of the interaction variable 'DMGEX' presented significantly negative at any significance levels. Therefore, we would say there were structural changes in export support financial policy. These structural changes may include the foreign pressures to the policy makers, openness and the shifts of the economic policy. This structural change in the export support financial policy influenced negatively on the change in the ESFP loans and on the growth rate of the ESFP loans. The government could no longer use the strong incentive parameter to give the reward of the domestic export support loans to the firms since the mid-1980s. The incentive parameter was lowered significantly and negatively due to structural changes. The government came to consider less the exports and more the other factors to determine the economic policy-makings. Once the government lowered the incentive parameter, the firm will react to the given policy to maximize its own objective function. We showed theoretically the firm would commit the less effort to the exports when the government lowered the incentive parameter. The political economy in those two different objective functions of the government and the firm made the firms to respond the different way from the socially desirable commitment. The fact that the heterogeneity of the interests among the government and the firms can R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 99 explain the firm's new choice of "less effort" under the moral hazard model at the next section. (b) Firm's commitment to exports in response to the incentive parameter change The next issue is whether the firms' commitment to exports is changed in relation to the incentive parameter changes. We just found there were structural breaks and negatively affect the export support loan decision of the government in the table 3-1 and 3-2. Here, we use the method of the Analysis of Variance to figure out whether there are differences of the firms' commitment to exports1 2 3 between before-structural break and after-structural break. We first divided the whole period into Group A and Group B. And then we tested if the growth rates of exports are significantly different depending on the groups. <Table 3-3> Scheffe's test for the growth rates of the exports: ANOVA Alpha = 0.1 Error Mean Square: 0.090057 Means with the same letter are not significantly different Scheffe Grouping Mean A 0.28544 Before the structural break B 0.12414 After the structural break From the result of Scheffe's test for growth rates of the exports in the ANOVA table, the firms' commitment to exports are significantly different depending on the 123. We used the growth rates of exports as the variables for the firms’ commitments to the exports between these two different periods. The differences between two different periods are evaluated at the 0.01 significance level with the Duncan’s Multiple Range test and Scheffe’s R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 100 periods which have two different incentive parameters. Firm's commitment to export growth was significantly decreased after the structural change since mid- 1980s. This result is to explain the role of heterogeneity of interests in decision makings. Second, we created the interaction dummy "DMGLO = DummyxGLO" to check how the structured changed ESFP loans affect the firms' export commitment in the regression. We expected that even after the structural breaks, ESFP loan was still important to the firm's decision, but the interaction dummy would negatively affect the firm's effort to export growth. The autocorrected regression model is as follows. <Table 3-4> Growth rate of Exports in response to new ESFP loan structure Dependent variable: GEX, the growth rates of Exports, R-square = 0.4414 Variable Parameter estimates t- value p-value Intercept 0.2194 5.57 <.0001 DM -0.0959 -1.84 0.0759 GLO 0.1979 2.18 0.0369 DMGLO -0.4386 -2.76 0.0097 1. Autocorrected for the serial correlation in errors by Maximum Like ihood Method Even after the structural break, the ESFP Loan is positively significant in the growth rate of exports1 2 4 , but the new structurally- changed ESFP Loan system (DMGLO) gives significantly negative effect in the export commitments. Firms would choose to decrease the exports commitments once they have unfavorable interaction loan dummy "DMGLO" to maximize their objective functions1 2 5 . multiple-comparison procedures. 124. The coefficient of the GLO (growth rate of the loan) is significant at 5 percent significance level. 125. Since the trade surplus (1985,86,87 and 88), the monetary authority had hard time in managing the monetary policy due to the money flow distortion and speculative money demand. Despite the fact that excess money demand was fallen and the ratio of the internal financing relative to the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 101 Therefore, with the analysis of the variance and the regression model, we would conclude the firms are responding negatively to the lowered incentive parameter changes after the structure change. 2. Contemporaneous Relationships and Simultaneity (a) Contemporaneous Relationships : only two variables ESFP Loans and Exports We will test whether there is a contemporaneous relationship between the first differenced export support loans and the amount of exports. The results are shown at the following table 3-5 and 3-6. Export support loans and exports affect significantly. Both tables are corrected for autocorrelation with the SAS autoreg procedures. <Table 3-51 2 6 > the first differenced Loans and Exports Dependent variable: DLO Durbin Watson Statistics Order DW Pr<dw HO 1 1.9285 0.3988 No serial correlation Variable Estimate T - value P - value equipment and inventory investment increased, the amount of external financing was increased a lot. This meant that the firms which earned export dollars, instead of investment in the equipment and repayment of the debts, seemed to purchase a lot of high return financial assets and land and buildings. <table> firms’ fund flow, unit: % __________________________________________________________ 1985 1986 1987 1988 Shortage in fimd/GNP 10.1 7.6 6.8 5.7 Internal Financing/Equipment & Inventory Investment 75.7 70.2 78.9 102.9 Land Purchase(Manufacturing)/Uses o f Funds 3.3 2.9 3.9 5.7 Uses o f Financial Assets/External Financing 13.7 23.2 32.8 52.4 1. Lee, Changkyu, The Theory o f Korea’s Financial System, p. 126 126. The table 4-10 and 4-11 show the result for the before-structural break period due to the structural changes since the mid-1980s. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 102 Intercept 470.6159 3.40 0.0011 DEX 0.3903 2.56 0.0126 <Table 3-6> the first differenced Exports and Loans Dependent variable: DEX Durbin Watson Statistics Order DW Corrected for the first order auto correlation HO 1 2.0292 N o serial correlation Variable Estimate T - value P - value Intercept 45.9187 1.11 0.2709 DLO 0.1912 4.17 <.0001 (b) Two Single Equations and Simultaneity (1) Two Single Equations: Relationships1 2 7 between ESFP Loans and Exports To investigate the relationships between the ESFP Loans and the Exports, we specified1 2 8 the ESFP loans in terms of the exports, the ESFP nominal interest rate, the interaction dummy variable for the all periods. The table 3-7 presents the nominal interest rate and real interest model and the table 3-8 presents what will happen if there is the spread of the two interest rates. Both ESFP real interest rate and nominal interest rate did not seem to have a significant influence on the ESFP loans, but the spread of the two interest rates influenced significantly at 10 percent level on the ESFP loans. 127. Due to the inclusion of the interaction dummy variable, this analysis is for the all periods. 1 2 8 . The rule of thumb is the parsimonious is better. However, the econometric model is better based on the economic theory which we set up at previous chapters. The specification of the two single equations came from the theoretical model. We tried other variables, but if the significance is so negligible, we dropped the other specifications. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 103 The equation is, GLO - a + fi0 GEX + fix DNR + y DMGEX (3-3) <Table 3-7> ESFP loans, Exports, ESFP Nominal Interest Rates, Interaction Dummy Dependent Variable: Gro\ R-square: 0.4454, AIC: 6.325722, vth Rates of ESFP Loans Vlaximum Likelihood Estimator Independent Variables Estimate t-value p-value Intercept 0.1691 1.98 0.0565 GEX 0.6118 2.08 0.0465 DNR1 2 9 0.0324 0.82 0.4174 DMGEX -2.1832 -4.13 0.0003 1. GEX: Growth Rates of Exports, DMGEX = Dummy* Growth Rate of Export, DNR: first difference in the Nominal Interest Rate. 2. The result is autocorrected for serial correlation in errors. The growth rate of exports had a significant impact on the growth rate of ESFP loans at 5 percent. Even in this specification, the interaction variable DMGEX is still negatively significant as we observed in the structural change analysis. The equation is , GLO = a + 0 o GEX + ftSO R + yDM GEX (3-4) <Table 3-8> ESFP loans, Exports, Spread of the two rates, Interaction Dummy Dependent Variable: Growth Rates of ESFP Loans R-square: 0.4985, AIC: 2.9037, Maximum Likelihood Estimator Independent Variables Estimate t-value p-value Intercept 0.1035 1.19 0.2439 GEX 0.2455 0.72 0.4746 SOR 0.0169 1.98 0.0568 DMGEX -1.6830 -3.02 0.0051 1 2 9 . When I use the Real Interest Rate: t-value is 0.25 and p-value is 0.8068. The conclusion for this interest rate is the same as the nominal interest rate. ESFP loan rate is not significant factor in the ESFP loans. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 104 1. GEX: Growth Rates of Exports, DMGEX = Dummy* Growth Rate of Export, SOR: General Interest Rate - ESFP Loan rate. 2. The result is autocorrected for serial correlation in errors. Table 3-8 continued. We would conclude, in this specification, the ESFP loan is influenced negatively by the interaction dummy, positively and significantly by the Exports, and the spread of the interest rates. (2) Two Single Equations: Relationships between Exports and ESFP Loans In the regression of the exports, we took the ESFP loans, Export Price Index, Wage Index, and the interaction dummy "DMGLO" as regressors. Based on the theory, we expect the wage index would have negative effect on the exports' commitment. The dollar term export price index shows negative coefficient but it does not significantly affect the growth rate of the export. The equation is, GEX = a + p fiL O + PD E P I+ /32DW I+yDM GLO (3-5) As we expected, the wage index gave a significantly negative impact on the growth rate of exports. The interaction dummy DMGLO influences negatively on the exports, too. <Table 3-9> The growth rates of Exports, Loans, Export Price Index, Wage Index, Interaction Dummy___________________________________________________ Dependent Variable: GEX, R-square: Likelihood 0.4758, AIC: -45.665903, Maximum istimator Independent variable Estimate t-value p-value Intercept 0.2560 5.47 <0001 GLO 0.1803 2.11 0.0440 DEPI -0.005338 -0.79 0.4341 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 105 DWI -0.0157 -2.49 0.0199 DMGLO -0.3737 -2.32 0.0274 Note: (1) GLO: Growth Rate of ESFP Loans, DEPI: the first difference in the Export Price Index (dollar base), DWI: the first difference in the wage index, DMGLO: the interaction dummy (2) The result is autocorrected for serial correlation in errors Table 3-9 continued. (3) Polynomial Distributed Lag Model in the Dynamic Economic Model: The model is this form. Y t = a + £ |3 i X h + U t (3-6), where i = 0 to K Almon lag assumes that pi can be approximated by a suitable degree polynomial in i, and the length of the lag. His insights are a smooth pattern of lag weights could be approximated by a polynomial of relatively low order. Almon technique has a distinct advantage over the Koyck method because the Koych has some serious autocorrelation problems that result from the presence of the stochastic explanatory variable and its correlation with the disturbance term. Once we use this Almon lag, we will figure out how the current and lagged independent variables influence the dependent variable with the smooth lag weights assumption. We will see the how lagged dependent export support loans (DLO_l) and the current and lagged exports1 3 0 as the explanatory variables influence the first differenced export support loans (DLO). <Table 3-10> Polynomial Distributed Lag model: Lag - 4, polynomial = 1_______ Dependent variable = DLO, the first difference export support loans 1 3 0 . The variables are DEX(O), DEX(l), DEX(3), and DEX(4) R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 106 Variable Estimates t-value p-value Intercept 259.06 1.40 0.1647 DLO 1 0.0410 0.33 0.7448 DEX**0 0.8249 2.47 0.0160 DEX**1 -0.1334 -0.41 0.6862 DEX(O) 0.453285 2.00 0.0499 DEX(l) 0.411102 2.55 0.0132 DEX(2) 0.368920 2.47 0.0160 DEX(3) 0.326737 1.63 0.1073 DEX(4) 0.284554 1.01 0.3164 Comment: DEX**0 and DEX**! are the estimates of the polynomials. Table 3-10 continued. From the above table, the current export, the lagged one of exports, and the lagged two of exports influence the government's export support loan amounts significantly at five percent significance level. The amount of export support loans are influenced within 6 months of the exports volume. Next issue is whether the first differenced exports are influenced by the lagged own variable, and the lagged export support loans. <Table 3-ll> Polynomial Distributed Lag model: Lag = 4, polynomial = 1 Dependent variable = DEX, t re first difference exports Variable Estimates t-value p-value Intercept 80.9 0.66 0.50 DEX_1 -0.6782 -6.68 <.0001 DLO**0 0.1558 2.06 0.0434 DLO**l -0.2011 -3.10 0.0029 DLO(O) 0.196831 3.70 0.0004 DLO(l) 0.133247 3.36 0.0013 DLO(2) 0.069663 2.06 0.0434 DLO(3) 0.006080 0.15 0.8783 DLO(4) -0.057504 -1.08 0.2832 Comment: DLO**0 and DLO**! are the estimates of the polynomials. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 107 The amounts of exports are influenced by the current and lagged export support loan variables significantly. (4) Simultaneity Issues and Hausman Specification Test Thus far, we considered only one-way or unidirectional cause and effect relationship. Such a one-way model emphasizes on the estimating and or predicting the average value of dependent variable conditional upon the fixed values of the explanatory variables. The cause-and-effect relationship, if any, in such models therefore ran from the explanatory to the explained. However, in many situations, such a unidirectional relationship is not so meaningful. This occurs if the regressands are determined by the regressors, and some of the regressors are, in turn, determined by the regressands. In short, there is a two way, or simultaneous, relationship between the regressors and the regressands, which makes the distinction between dependent and explanatory variables of dubious value. One of the crucial assumptions of the method of OLS is that the explanatory variables are either nonstaochasitic or, if stochastic (random), are distributed independently of the stochastic disturbance term. If neither of these conditions is met, then the least-squares1 3 1 estimators are not only biased but also inconsistent. 1 3 1 . When the OLS method is applied with disregarding other equations in the system. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 108 This suggests that we should check for the simultaneity problem before we discard OLS in favor of the alternatives. A test of simultaneity is essentially a test of whether (an endogenous) regressor is correlated with the error term. To find out which is the case in a concrete situation, we can use Hausman's specification error test. In the structural equations1 3 2 in a multi-system, we already considered the following specifications with maximum likelihood estimation method in the single equation estimation methods in the above. G LO - a q + a^GEX + a 2 DRR + a ^ D M G E X + ul( (3 - 7) GEX = /3 0 + PxGLO + p 2 DEPI + p ^ D W I + p 4DMGLO + u 2{ (3 - 8) A Version of the Hausman specification error test that can be used for testing the simultaneity problem can be explained as follows for the two equations (3-7) and (3- 8). Assume that the DRR, DEPI, and DWI are exogenous. Of course, GLO and GEX are endogenous. If there is no simultaneity problem, GLO and Lht should be uncorrelated. On the other hand, if there is simultaneity, GLO and Lta will be correlated. To find out which is the case, the Hausman test proceeds as follows. First, from the above two equations, we obtain the reduced-form equations which are composed of predetermined variables. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 109 GLO, = n 0 + n, DRR, + n 2 DEPI, + n 3 D WI, + V, (3-9) G E X , = Yl,+ YI 5 DRR,+ n 6 DEPI, + Y\n DWI, + W, ( 3 - 1 0 ) where V and W are the reduced form error terms which are white noise. Estimating (3-9) by OLS we obtain the predicted value GLOHAT for GLO and LVHAT for Vt. Substituting these GLOHAT and LVHAT for equation (3-8), we get GEX = /?0 + p xGLOHAT + LVHAT + P2DEPI( + P1 )DWIt + P4DMGLO( + u2( (3-11) , where the coefficient of GLOHAT and LVHAT are the same. Now under the null hypothesis that there is no simultaneity, the correlation between Vt and Wt should be zero, asymptotically. If we run the regression equation (3-11), and find that the coefficient of LVHAT is statistically zero, we can conclude that there is no simultaneity problem. <Table 3-12> Hausman specification test of GLOHAT and LVHAT Dependent Variable: GEX, R-square: 0.4760, AIC: -43.68 Independent Variable Estimate T-value P-value Intercept 0.2188 0.54 0.5923 GLOHAT 0.2832 0.25 0.8007 LVHAT 0.1800 2.07 0.0483 DEPI -0.004087 -0.27 0.7888 DWI -0.0122 -0.32 0.7551 DMGLO -0.3741 -2.28 0.0301 132. One of the important issues for the estimation of those two equations is Identification Issues as well as the Simultaneity. By the Order Condition of Identifiability, these two equations are over identified. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 110 The coefficient of LVHAT is statistically different from zero at 5 and 10 percent significance level, implying that we can reject the no-simultaneity null hypothesis. The other way for the GLO equation showed the similar result1 3 3 of rejecting the no simultaneity null hypothesis. D. Investigating the inter-relationships in a Vector Autoregression (VAR) Model 1. Issues and Theory of Standard VAR(P) MODEL In the above, we considered simultaneous equation models. In such models some variables are treated as endogenous and some as exogenous or predetermined. Before we estimate such models, we have to make sure that the equations in the system are identified. This identification is often achieved by assuming that some of the predetermined variables are present only in some equations. This decision is often subjective and has been severely criticized by Christopher Sims1 3 4 . Traditionally, empirical research work (regardless of macro- or microeconomics) begins with the use of theory to construct a highly restricted structural econometric model. The determination of the nature of those restrictions is fraught with difficulty and thus many specification decisions turn out to be largely ad hoc and of 133. Estimated Residual EVHAT = GEX-GEXHAT shows the significance at 10 percent significance level. 134. Christopher Sims, 1980, op,cit,p3. : “what economic theory tells us about them is mainly that any variable which appears on the right-hand side of all of them. To the extent that models end up with very different sets of variables on the right-hand-sides of these equations, they do so not by invoking economic theory, but by invoking an intuitive, econometrician’s version of psychological and sociological theory, since constraining utility functions is what is involved here. Furthermore, unless these sets of equations are considered as a system in process of specification, the behavioral implications of the restrictions on all equations taken together may be much less reasonable than the restrictions on any one equation taken by itself.” R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. I l l dubious validity1 3 5 . Once these specification issues have been addressed and a structural model constructed, it is then estimated using complicated econometric techniques. This is what we have been doing so far. An alternative approach to modeling economic time series is the Vector Autoregression (VAR) model introduced by Sims (1980). This VAR is to analyze time series relationships among economic variables. The fact that the use of this approach has proliferated widely in a relatively brief period of time is due not only to the very persuasive criticisms of traditional methods made by Sims1 3 6 , but perhaps even more to the perception that specification and estimation problems are greatly simplified. The benefit of the VAR approach imposes few restrictions and can generally use OLS estimation procedures. The VAR approach is thus believed to be largely free of the spurious specification assumptions and consequent specification errors necessitated by traditional econometric procedures. Sim's1 3 7 (1980) criticisms of the "incredible identification restrictions" inherent in structural models argue for an alternative estimation strategy. By the Wold decomposition theorem, a covariance stationary vector time series process has a vector moving average representation, that is, each variable can be 135. Spencer, David, “Does money matter? The Robustness of Evidence from Vector Autoregressions”, Journal of Money, Credit and Banking, Vol.21, No.4, pp. 442-443 1 3 6 . Sims, Christopher, “Macroeconomics and Reality”, Econometrica 48 (Jan. 1980), 1-49. 1 3 7 . Sims (1980), op cit. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 112 expressed as a linear combination of its own current innovation1 3 8 (errors) and lagged innovations of all the variables in the system. For example, let7 s think about the first order VAR model. X t = A 0 + A }X ,_! + et (3-11) where et ~N(0, X) and E(etes ') = 0n for all s t . After n iterations, xt - (i+A + — ■ ■ + A")A>+ X 4 e< -‘ + < = 0 If we continue to iterate backward, it is clear that convergence requires that the expression A" vanish as n approaches infinity. Assuming the stability condition is met, we can write the particular solution for xt as Vector Moving Average (VMA) form, o o xt =V + Y s A1 V 1 (3-12) ;= o where /u is a mean of vector x( and et ~N(0, £) and E(etes ’ ) = 0n for all s * t .. In general, however, the innovations are contemporaneously correlated1 3 9 and so a unique decomposition does not exist. This problem is usually resolved by applying a triangular orfhogonalizing transformation1 4 0 "Choleski factorization of 138. Enders, op cit. pp.264-268 139. Due to the contemporaneous relationships between the variables, the disturbance terms are composites of pine innovations. Therefore, when calculating Impulse Response functions and Forecast Error Variance Decompositions, the answer to the question that how an innovation to variable i affect the variable j is not available. 1 4 0 . Following Lutkepohl(p.9-49, 1997), to see how the innovations in one endogenous variable affects another, we need to consider a VAR(p) that does not include exogenous variables, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 113 the positive definite estimated covariance matrix1 4 1 ." This transformation is not unique and by selecting one, the researcher essentially arbitrarily imposes a particular causal ordering on the variables in the VAR model. If the moving average representation mentioned above is expressed in terms of these orthogonalized innovations the forecast errors for each time series variable As discussed, it can be rewritten as an infinite order moving average process; we can write y t as the sum of infinite series of serially uncorrelated innovations but contemporaneously correlated innovations. Specifically, if the VAR(p) is stable, it can be written as The ® s are known as the simple impulse-response functions at horizon s. The i,j element of O v give s the effect of a one-time unit increase in an innovation to variable j on variable i after s periods holding everything else constant. As will be discussed in the impulse response function section, the problem is that since E[utut '] is not restricted to being a diagonal matrix, an increase in an innovation to one variable provides information about the innovations to other variables. This implies that no causal interpretation of the simple impulse response function is possible. However, suppose that we had a matrix P such that Z = PP' • It can be shown that the variable in p 'w ( have zero mean and that E{PAut (P )'} = IK. We rewrite VMA as If we had such a P, then the WK would be mutually orthogonal innovations, and no information would be lost in the ceteris paribus assumption. The © v would have the causal interpretation that we seek. Then, the question is how to find a P in a VAR(p) model. We follow Sims(1980) Cholesky Decomposition of Z to identify the structural equations with P. 141. Johnston, Econometric Methods, 4th edition, pp.287-301 yt =v + Aly,_l +..... + Apy t_p + ut , where ut ~(0,Z) y t = n + O v «(v where // is the K x 1 time-invariant mean of y t and s= 0 IK if S = 0 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 114 can be decomposed uniquely (for that orthogonalization) into the sum of contributions from each of the innovations in the system. This allocation of forecast error variance attributed to each particular innovation is called Variance Decomposition. The following four sections are committed to the issues in a VAR estimation. I consider these four issues in considering VAR(p) model estimation. (a) Alternative Specifications1 4 2 Although VAR methods are often referred to as "atheoreticaF, there are numerous specification issues that would necessarily be addressed beyond the choice of variables to be included. First issue is the lag length. Estimation of a VAR model requires the explicit choice of lag length in the equations of the model. Alternative choices will give different innovation series and thus will likely make a difference in the variance decomposition results. Deciding on the maximum lag length is an empirical question. In the annual data, we have 35 observations from 1962 to 1996 and in the quarterly data, we have 124 observations from 1967 to 1997. Including too many lagged terms will consume degrees of freedom, not to mention introducing the possibility of multicollinearity. Including too few lags will lead to specification errors. One way1 4 3 of deciding this question is to use a criterion 1 4 2 . Spencer, David, op. cit, pp.445-446 143. Or testing for appropriate lag length using a standard F test. Given that, it would be seem to be important to be appropriately generous in practice. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 115 like the Akaike or Schwarz and choose that model that gives the lowest values of these criteria. There is no question that some trial and error is inevitable. In practice these lags have typically been rather short, a year or less (four quarters, twelve months, etc., depending on the frequency of observation), although Sims (1980) estimates a VAR with both one-and two-year lags. At later section, we would see the Akaike Information Criterion on the lags. (b) Ordering of Variance The conventional orthogonalization procedure requires imposing a particular causal ordering of the variables. This choice is arbitrary and, when there is contemporaneous correlation among the innovations, it can make a significant difference for the variance decomposition. Innovation accounting analyses are often conducted on a system with orthogononalized innovations1 4 4 . The most commonly used orthogonalization requires the selection of an ordering of the variables in the system, since it attributes responsibility for contemporaneous correlation between any two variables' innovations to the variable that is higher in the ordering. Thus an innovation in the first variable in the ordering is assumed to influence all variables contemporaneously without being influenced by any of the other variables, 1 4 4 . The innovations in the VAR are, in general, not contemporaneously independent of one another. That one innovation receives a perturbation and the other does not is implausible. A widely used solution to the problem is to transform the innovations in the standard VAR to produce a new set of orthogonal innovations. These will be pariwise uncorrelated have unit variances. Johnston, Econometric Methods, pp.298-301, Fourth Edition. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 116 and, at the other extreme, an innovation in the last variable influences only itself contemporaneously.1 4 5 This problem has been noted by Sims and widely criticized as a deficiency of VAR methods by others1 4 6 . 145. It is as if the VAR was re-estimated with current observations on the variables entering via a lower block triangular matrix of coefficients. Thus the first variable in the ordering has no current values on the right hand side of its equation, the second variable includes the current value of only the first variable, and so on, down to the equation for the last variable in the ordering, which contains the current values of all the other variables in the system (Sims, 1980, pp.21-23, Recursive system proposed). Decomposing the residuals in the lower triangular fashion influences the variance decomposition values different from the upper triangular ones. For example, from the structural bi- variate system: called structured VAR model, yt = bm - bn z, + yny t + yn zt_ x + z t = ^ 2o — ^21 + T^iTr-i ^~Y nz t-\ + S zt Estimating the system using OLS yields the theoretical parameter estimates after imposing the restrictions: yt =alo+auy^+anz,-i+eu Z t = a 20 + a 2 l T f - l + a 22Z t - l + e 2t The lowering triangular Choleski Decompositon is, \ e2 t ) ( 1 0 -b2l 1 So ^ 'yt \ G zt j 'yt \ s zt b 2x j The upper triangular Choleski Decomposition is, \ e 2t J 1 - b , 12 U \ 'yt \ £ zt j ' yt 12 V £zl J In the upper triangular fashion, both €yt and £n shocks affect the contemporaneous value of y t , but only £zt shocks affect the contemporaneous value of z t . The observed values o f e2l are completely attributed to pure shocks to the z t sequence. Therefore, the numerical variance decompositions are often very sensitive to the order in which the original innovations are orthogonalized. Enders, Walter. Applied Econometric Time Series, Second edition, pp.264-272 1 4 6 . Gordon and King(1982), Cooley and LeRoy (1985), and Learner (1985). James D. Hamilton, Time Series Analysis, 1994, pp. 324-336 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 117 To resolve this potential sensitivity, as one way to follow Doan and Litterman1 4 7 (1986), an attempt can be made to place the variables in order of decreasing exogeneity from an a priori perspective. As another way, similarly, following Gordon and King1 4 8 (1982), the variables which is thought to be most sensitive to current innovations should be placed low in the ordering; equivalently, for example, putting Exports after the ESFP loans means we assume it is more likely that an innovation in ESFP loans will lead to an innovation in Exports contemporaneously than that an innovation in Exports will lead to a contemporaneous innovation in production and vice versa. Therefore, it is worthy of trying a few different orderings to see the different variance decomposition values depending on the orderings to investigate the inter relationships among the variables such as exports, export loans, and interest rates, (c) Trend Removal The validity of the VAR approaches relies on the presumption that the economic variables under considerations are covariance stationary. Thus, it may be important to induce stationary time series by appropriately transforming1 4 9 any nonstationary series referred as "trend removal". 1 4 7 . Doan and Litterman (1986), RATS User’s manual version 2.0, VAR Econometrics. They suggested that the national variables be placed ahead of the sectoral variables. 1 4 8 . Gordon, R.J. and S.R. King (1982), ‘The output cost of disinflation in traditional and vector autogressive models,’ Brookings Paper on Economic Activity, 205-42. 149. Sims (1980), Litterman and Weiss (1985) using VAR methods to investigate the money-output relationship include no trend correction. R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 118 However, there is an issue of whether the variables in a VAR need to be stationary. Sims (1980) and Sims, Stock, and Watson (1990)1 5 0 recommend against differencing even if the variables contain a unit root. They argue that the goal of a VAR analysis is to determine the interrelationships among the variables, not to determine the parameter estimates. The main argument against differencing is that it "throws away" information concerning the co-movements in the data (such as the possibility of cointegrating relationships). Similarly, it is argued that the data need not be detrended. In a VAR, a majority view is that the form of the variables in the VAR should mimic the true data generating process. This is particularly true if the aim is to estimate a structural model. In this paper, we checked the stationarity of all of variables and found out most of time series in this data have unit root which had to be first differenced in log form. We follow the majority view that all of the variables are transformed by differencing to be stationary. (d) Level of Temporal Aggregation The fact that contemporaneous correlation among innovations1 5 1 is likely to increase as the data become more temporally aggregated suggests that results using semiannual or annual data are likely to be more sensitive to choice of ordering. 1 5 0 . “Inference in Linear Time Series Models with some Unit Roots”, Econometrica, vol 58, 1990, pp.l 13-144. 151. Disturbance terms in reduced from VAR(p) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 119 Furthermore, using temporally disaggregated data increases the sample size. This would seem to imply a preference for monthly or quarterly data, but there are several reasons why we might expect the use of annual data to be more useful in correctly capturing the inter-relationships between the growth rate of exports and the growth rate of ESFP loans. If the data is too frequent, it may be hard to reflect the natural interval in the relationships among the variables. Additionally, disaggregated data may contain significant random measurement error and thus have a high proportion of noise making it difficult to empirically discover relationships which in fact may exist. Data which have been aggregated by averaging over time are likely to be less noisy and thus may be better able to correctly find such relationships. However, we examine the variance decomposition results only for quarterly data because of the limited number yearly data observations and the degree of freedom of the annual data with the p-th1 5 2 order VAR. 2. A Estimation of a Standard VAR(p) In principle, there is nothing to prevent us from incorporating a large number of variables in the VAR. It is possible to construct an n-equation VAR with each equation containing p lags of all n variables in the system. I include those variables 152. Based on Akaike Information criterion, I will give 4 lags in a VAR estimation. That means with annual data three variable equations, degree of freedom consumed a lot of observations. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 120 that have important economic effects on each other. However, as a practical matter, degrees of freedom are quickly eroded as more variables are included. The data starts from 1967.1 to 1997.4. which has the 124 observations of quarterly data with three variables - Exports, Export Support Loans and the ESFP loan rate or Spread 1 5 3 of the interest rates. Due to the problem of degree of freedom in annual data, for a VAR(p), I used quarterly data to do VAR(p) analysis. The variables used in this reduced form VAR were ESFP loans, Export1 5 4 , and the spread of interest rate. A 3-equation VAR can be represented by ( A \ "10 ' a u (L) Al2(L) A A i y ( x An i '*u' X2I = A2o + A2l(L) a 22(L) A23(L) X2l-l + e2 i <X3I ; <^30 , U ,(£> a 32(L) A33( L ) / VX3(-1 \ e31 y where: Aio= the parameters representing intercept terms Aij (L) = the polynomials in the lag operator L Xt = { Export, ESFP Loans, SOR}' The individual coefficients of Aij(L) are denoted by aij(l), aij(2),..... Since all equations have the same lag length, the polynomials A ij(L ) are all of the same degree. 153. SOR = spread of interest rate = General loan rate - ESFP loan rate 1 5 4 . ESFP loan and Exports are the first difference in log which is an approximation of growth rate. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 121 The terms ey are white-noise disturbances that may be correlated across equations. Let's designate the variance-covariance matrix by £ , where the dimension of X is 3 by 31 5 5 . In addition to the determination of the set of variables to include in the VAR, it is important to determine the appropriate lag length. One possible procedure is to allow for different lag lengths for each variable in each equation. However, in order to preserve the symmetry of the system1 5 6 , it is common to use the same lag length for all equation. OLS estimation will give consistent and asymptotically efficient estimates. If some of the VAR equations have regressors not included in the others, seemingly unrelated regressions (SUR)1 5 7 provide efficient estimates of the VAR coefficients. In a VAR, long lag lengths quickly consume degrees of freedom. If the lag length is p, each of the n-equations contains np coefficients plus the intercept term. Appropriate lag length selection can be critical. If p is too small, there may be a chance that the model is misspecified. If p is too large, degrees of freedom are wasted. To check lag length, begin with the longest plausible length or the longest feasible length given degrees-of-freedom considerations. < 7 ,2 C / i C / I M I I cr2 i °22 °23 ^31 °32 °33 , 156. Therefore, we are able to use the OLS. 157. Hence, when there is a good reason to let lag lengths differ across equations, estimate the so- called near-VAR using SUR model. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 122 Next we estimate the VAR and form the Variance-covariance matrix of the residuals. Using the quarterly data observation with three variables, I start the lag length 4 quarters based on a priori notion that one year is sufficiently long1 5 8 to capture the system's dynamics. Then I presented FPE1 5 9 (Final Prediction Error), AIC(Akaike Information Criterion), HQIC(Hannan-Quinn Criterion) and SBC(Schwarz Bayesian Criterion) in each different lag to select the optimal lag order. All1 6 0 of the criterion are based on Lutkepohl (1993, p.l25-135)'s methods. All of the information criterion1 6 1 has the smallest value when we have a lag order "4", in the quarterly data, in the following table 3-12 depending on the different ordering, except that the SBC shows the first order lag has the lowest criterion. <Table 3-12-l> VAR(p) order selection criteria_________________________________ ____________________ In the ordering: Loan Export SOR____________________ 158. Based on the economic model, the impact on the endogenous variables is to happen within one year. 159. Lutkepohl (p. 128): the FPE is really not an information criteria. As the name suggests the FPE has its origins as a measure of forecast inaccuracy. Thus, it is intuitive that the VAR(p) with the lowest FPE is, by some measure, optimal. The FPE is quite similar to AIC. Conceptually, the AIC is designed to measure the discrepancy between the given model and the true model, which we would want to minimize. As Ltitkepohl (p.130-132), choosing p to minimize the SBIC and HQIC provides consistent estimators of the true lag order p. That is to say, whether an estimator has a desirable asymptotic property such as p lim p = p or, equivalently, lim Pr {p = p ) = 1 r->. 160. Lutkepohl (p. 133): In small samples AIC and FPE may have better properties than HQ and SBIC. The former two criteria are designed for minimizing the forecast error variance. Thus, in small as well as large samples, models based on AIC and FPE may produce superior forecasts although they may not estimate the orders correctly. In fact, Shibata (1980) derives asymptotic optimality properties of AIC and FPE for univariate processes. He shows that asymptotically they indeed minimize the 1- step forecast MSE. 161 R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 123 P=1 P=2 P=3 P=4 Final Prediction Error Criterion (FPE) 0.0001114 0.0001184 0.000105 0.0000774* Akaike Information Criterion (AIC) -9.15253 -9.09188 -9.21353 -9.51961* Hannan-Quinn Criterion (HQ) -9.06718 -8.92118 -8.95748 -9.17821* Schwarz Bayesian Criterion -8.94234* -8.67151 -8.58297 -8.67887 Note: (1) A superscript * appears next to the smallest statistic indicating that the corresponding lag is optimal. (2) SOR means that spread of interest rate between the general rate and ESFP loan rate. Table 3-12-1 continued. Even in the ordering for the Export to be ahead of the ESFP loan, all of the information criterion shows the very similar result to the table 3-12. VAR(4) estimates show the following results. Endogenous variables such as ESFP loan, Export, and SOR are determined by the predetermined variables with a constant term at table 3-13. This VAR(4) has a constant term in each equation. In the ESFP loan equation, Export.L3 is significant at 5 percent significance level in determining ESFP loans. Other Export lags are not significantly influencing ESFP loans. In the equation of Export, ESFP loan of lag 1 influences positively and significantly the exports at 10 percent significance level. Most of loan lags do not affect significantly exports at any significance level. <Table 3-13> VAR(4) estimation Equation Coefficients t-value P>ltl R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 124 Loan.Ll 0.4329193 4.46 0.000 Loan.L2 -0.0131193 -0.12 0.902 Loan.L3 -0.0328356 -0.31 0.755 Loan.L4 0.028727 0.31 0.760 Export.Ll -0.0844795 -1.20 0.231 Export.L2 -0.0216479 -0.31 0.761 ESFP loan Export.L3 0.140685 2.04 0.044 Export.L4 -0.0124808 -0.18 0.858 SOR.L1 -0.0036397 -0.50 0.615 SOR.L2 0.0026583 0.26 0.799 SOR.L3 0.0058557 0.56 0.576 SOR.L4 -0.0009751 -0.13 0.893 Constant -0.0056877 -0.44 0.664 Loan.Ll 0.1909942 1.77 0.079 Loan.L2 -0.0985735 -0.84 0.404 Loan.L3 -0.168934 -1.45 0.151 Loan.L4 0.1439679 1.38 0.169 Export.Ll -0.1942552 -2.50 0.014 Export.L2 0.0056486 0.07 0.945 Export Export.L3 -0.1565404 -2.04 0.044 Export.L4 0.5478423 7.08 0.000 SOR.L1 0.006775 0.85 0.400 SOR.L2 -0.0100014 -0.87 0.389 SOR.L3 0.0056897 0.49 0.624 SOR.L4 0.0001383 0.02 0.986 Constant 0.0178136 1.23 0.221 Loan.Ll -1.177075 -0.91 0.367 Loan.L2 0.0709801 0.05 0.960 Loan.L3 1.383299 0.98 0.327 Loan.L4 -1.025932 -0.82 0.415 Export.Ll -0.915613 -0.98 0.331 Export.L2 -0.5497706 -0.58 0.563 SOR Export.L3 0.105903 0.11 0.909 Export.L4 0.1667027 0.18 0.858 SOR.L1 1.042236 10.79 0.000 SOR.L2 -0.0840756 -0.60 0.547 SOR.L3 0.0813838 0.58 0.561 SOR.L4 -0.0666433 -0.69 0.492 Constant 0.1288808 0.74 0.461 Note: 1. variable.L# = lag # of each variable 2. Even Without a constant term , similar results are estimated. Table 3-13 continued. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 125 In an equation of SOR, both export and loan do not affect significantly and SOR is influenced by its first lag. From this a VAR(4), we see the export loan decision by government is positively influenced by the export with lag 31 6 2 and the export by firm is influenced by export support loan of government over the overall period. 3. A standard VAR, Granger Causality Test with foreign pressures and export price index One of the key questions that can be addressed with vector autoregressions is how useful some variables are for forecasting others. Granger1 6 3 's idea is that a cause cannot come after the effect1 6 4 . More generally, since the future cannot predict the past, if variable X Granger causes variable Y, then changes in X should precede changes in Y. Therefore, in a regression of Y on other variables (including its own past values) if we include past or lagged values of X and it significantly1 6 5 improves the prediction of Y, then we can say that X Granger causes1 6 6 Y. 1 6 2 . This makes sense when government used the past export accomplishment for the loan decision. 163. Granger, C.WJ. (1969), “Investigating Causal Relations by Econometric Models and Cross- Spectral Methods,” Econometrica, 37, 424-438 1 6 4 . Gary Koop, Analysis o f Economic Data, John Wiley & Sons, New York, 2000, p.175: For the dependence of one variable on other variables in time series regression, Koop wrote, “ time does not run backward. That is, if event A happens before event B, then it is possible that A is causing B. However, it is not possible that B is causing A. In other words, event sin the past can cause events to happen today. Future events cannot.” 165. To test the Granger causality, we apply the F test, F (RSSr -R S S ur)/m RSSu r /( n - k ) 1 6 6 . The econometrician Edward Learner prefers the term “ precedence” over causality. Francis Diebold prefers the term “predictive causality” (Elements of Forecasting, South Western Publishing, 2001, p.254). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 126 Based on the idea of Granger causality, I will investigate a vector autoregression and its Granger causality of a model that has two endogenous variables1 6 7 and one exogenous1 6 8 variable which is Foreign Pressure such as international financial institutions and US government pressure. When there are foreign pressures against the export support loan policy, all of the pressure dummy in each equation shows statistically significant. Foreign pressure influences negatively the export support loan and exports in loan and export VAR equations. Granger causality analysis shows that export Granger causes the export support loan, but the loan does not Granger causes the export when there is a foreign pressure exogenous variable. The result is shown in the table 3-15. The Second VAR model has the loan, export, and export price index and Granger causality is investigated. The result of VAR(4) estimation is shown at table 3-16. In the loan equation, export of lag3 influences significantly positively the export support loan. Export price index does not affect significantly the loan. 167. Export support loans and exports in the log differences. 168. Exogenous variable here is a nonstochastic variable that is a dummy for the foreign pressure for the equation. The model is : P s y, + X 0 < x<- +st ;= i i= i where yt = (export support loans, exports) and x, = (foreign pressure dummy) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 127 Actually and controversially Granger causality is affected by the lags and what other variables are included in the VAR. It may be used as one of the reference to observe the inter-relationships among variables due to various levels of criticisms1 6 9 . <Table 3-14> VAR(4) with a foreign pressure exogenous variable Equation Coefficients T-value P>ltl Loan.Ll 0.438399 4.57 0.000 Loan.L2 -0.0116218 -0.11 0.912 Loan.L3 -0.0236896 -0.23 0.820 Loan.L4 0.0393708 0.43 0.666 ExportLl -0.0812137 -1.17 0.244 ESFP loan Export.L2 -0.0106075 -0.15 0.800 Export.L3 0.1498453 2.20 0.030 Export.L4 -0.0115982 -0.17 0.867 Foreign Pressure -0.0393582 -2.11 0.037 Constant 0.0374101 2.08 0.040 Loan.Ll 0.2058875 1.94 0.055 Loan.L2 -0.1125766 -0.97 0.334 Loan.L3 -0.1579584 -1.37 0.173 Loan.L4 0.1611955 1.60 0.113 Export.Ll -0.1889649 -2.46 0.015 Export Export.L2 0.0070295 0.09 0.928 Export.L3 -0.15062 -1.99 0.049 Export.L4 0.5509011 7.22 0.000 Foreign Pressure -0.214264 -1.04 0.302 Constant 0.0429317 2.16 0.033 <Table 3-15> Granger causality of export support loans and export when there is a foreign pressure________________________________________________________ Granger Causality test In the Equations Excluded F value Prob > F Export Support Loans Export 2.1621 0.0780 Export Export support loan 1.8075 0.1326 Table 3-15 continued. 169. See. Lutkepohl (1997), Introduction to Multiple Time Series Analysis, p35-43. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. In the table 3-16, In the export equation, the loan of lag 1 and lag 4 influences significantly the export support loans. As expected, the export price index of lag 4 influences significantly negatively the exports. In the export equation, the export price index shows significant and negative sign on the exports at lag 4. We would guess it takes a time for the price index to affect the exports. <Table 3-16> Vector Autoregression: Loan, Export, Export Price Index Equation Coefficients t-value P>ltl Loan.Ll 0.5223888 4.95 0.000 Loan.L2 -0.0407496 -0.34 0.732 Loan.L3 0.0667383 0.56 0.577 Loan.L4 0.0310413 0.29 0.773 Export.Ll -0.0118327 -0.14 0.890 Export.L2 0.0357509 0.41 0.681 ESFP loan Export.L3 0.1620441 1.94 0.056 Export.L4 0.0011123 0.01 0.989 EPI.L1 0.3196348 0.48 0.629 EPI.L2 0.1166863 0.14 0.890 EPI.L3 -1.282654 -1.53 0.130 EPI.L4 0.7495741 0.0051846 1.12 0.42 0.266 0.677 Loan.Ll 0.1855058 1.84 0.070 Loan.L2 -0.1109107 -0.98 0.332 Loan.L3 -0.1946911 -1.70 0.092 Loan.L4 0.2483533 2.42 0.018 Export.Ll -0.239302 -2.93 0.004 Export Export.L2 -0.0220984 -0.27 0.791 Export.L3 -0.1354479 -1.69 0.094 Export.L4 0.573649 7.31 0.000 EPI.L1 1.770861 2.81 0.006 EPI.L2 0.1358013 0.17 0.866 EPI.L3 -0.1815023 -0.23 0.822 EPI.L4 -1.952451 -3.05 0.003 Loan.Ll 0.0026562 0.16 0.874 EPI Loan.L2 -0.0200527 -1.07 0.289 Loan.L3 0.0216717 1.15 0.254 Loan.L4 -0.0120946 -0.71 0.478 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 129 ExportLl Export.L2 Export.L3 Export.L4 -0.0032931 0.0064832 0.0109916 -0.0091581 -0.24 0.47 0.83 -0.71 0.808 0.638 0.408 0.482 EPI.L1 0.8023498 7.70 0.000 EPI.L2 0.0659448 0.50 0.620 EPI.L3 0.0021991 0.02 0.987 EPI.L4 -0.211947 -2.00 0.048 (1) EPI: the first difference in the export price index. Table 3-16 continued When there is export price index included in the VAR(4), the Granger causality1 7 0 shows that the loan Granger causes the exports and the price index Granger causes the export, too. 4. Structural Vector Autoregression: Forecast Error Variance Decomposition (a) SVAR: Identification and Choleski Decomposition1 7 1 Sims(1980, 82) suggested scrapping simultaneous equation systems altogether, and to use models whose specification had to be founded on the analysis of the statistical properties of the data under study. In fact, Sims suggested to specify VARs, i.e. multivariate models where each series under study is regressed on a finite number of lags of all the series jointly considered. Clearly, in a VAR model contemporaneous relationships among variables are not accounted for and are "hidden" in the instantaneous correlation structure of the error terms. 170. Granger causality Tests Equation Excluded F P-value Export Loan 2.6248 0.0398 Export EPI 7.2789 0.0000 171. Amisano and Giannini, Topics in Structural VAR econometrics, pp. 1-28, 1997, Springer, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 130 However, main conceptual problem in the VAR use is related to the interpretation of the instantaneous correlations among error terms, and therefore among observable variables. Structural VAR is based on the attempt to give a sensible solution to this problem, based on the imposition of a set of restrictions. One way of describing an SVAR is to illustrate how it links the VAR framework and the dynamic structural simultaneous equation framework. One way of imposing the Cholesky1 7 2 restrictions on this VAR system is by applying equality constraints using the constraint matrices. The triangular Sims’(1980) original proposal consisted in moving from a non-orthogonal VMA to an orthogonalized VMA representation via Choleski factorization of the £ matrix. This amounts to starting from the reduced form VAR representation A(L)y, - s t, s r VWN{0 ,1 ) and to pre-multiply the system by the inverse of the Choleski factor of £ A\L)y,= e„ e,~VWN(0,I„) 4 = p ' , a ; = p - '4 (=0 P P ' = I where P is the Choleski factor of X, and clearly is lower triangular with unit diagonal elements. This amounts to modeling contemporaneous relationships among the endogenous variables in a triangular recursive form. The resulting orthogonal VMA representation is y ,= lc ,P e ,_ = ± < S > ,e ,_ , (=0 (=0 ® ,= C /5 ®0 = P Notice that, since ® 0 = P , the orthogonal VMA representation shocks et have instantaneous effects on the elements of y t according to the triangular scheme given by the Choleski factor P. R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 131 representation, which is sometimes referred to as Wold causal chain, is clearly a very particular one which cannot be considered suitable to every applied context. Sometimes the researcher might have in mind different constraints for formalizing the contemporaneous relationships between the endogenous variables. Therefore, based on the economic theory presented here, we assume that the change in ESFP loan is not contemporaneously affected by the changes in either export or spread of interest rates. We also assume that exports are NOT affected by contemporaneous changes in ESFP loan due to the sequential decisions of response function of firm to government loan decision. Export is influenced by a change in innovations to ESFP Loan with lag 1. It is assumed that the spread of interest rate is affected contemporaneously by ESFP loan and exports to see whether the rate is contemporaneously subject to the market. In the economic theory presented in the model section, the government announced the reward structure and sequentially expects the reaction function from firms. The decisions of firm and government are made sequentially and markets Moreover, it is true that given the matrix £ , the Choleski factor P is uniquely determined. Nevertheless, if the elements of y t were permuted and arranged in yt , the rows and columns of £ would have to be permuted accordingly to generate • The matrix X would then have a different Choleski factor: p * p*< — which would produce a different orthogonalized VMA representation. Therefore, the orthogonal VMA representation corresponding to the Choleski decomposition of variance-covariance matrix of the reduced form disturbances is unique only given a particular ordering of the observable variables contained in y t . R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 132 are reacting contemporaneously to those decisions of firm and government. Instead of imposing Cholesky Decompositions on SVAR to be identified, according to the moral hazard model presented here, let7 s give a first try of such restrictions on A and B constraint matrix. . We first use the constrainted AB1 7 3 matrix form of Amisano and Giannini (1997) to structuralize VAR model. Constraints of matrix1 7 4 are the followings according to the theory. 173. Amisano and Giannini, Topics in Structural VAR econometrics, 1997, second revised version. p.15-19 A, B are n x n invertible matrices such that: A A(L)y, = A s, Ast = Be, E(e,) = 0 E(e,e,') = In In this kind of structural model, it is possible to model explicitly the instantaneous links among the endogenous variables, and the impact effect of the orthonormal random shocks hitting the system. Notice that the A matrix induces a transformation on the s, disturbance vector, generating a new vector ( A s ,) that can be conceived as being generated by linear combinations (through the B matrix) of n-independent (orthonormal) disturbances, which we will refer to as e ,. Obviously this structure might have a different menaing than those of models K and C. Notice that the AB-model can be seen as the most general parameterization nesting the C and K models special cases. In fact, the C-model can be seen as a particular case of the AB-model, where A is chosen to be identity matrix, and the K-model corresponds to an AB-model with a diagonal B matrix. As in the previous case, from As, = Be, As,s, 'A ' = BB' for £ known, this equation again imposes a set of n(n+l)/2 non-linear restrictions on the parameters of the A and B matrices, leaving overall 2n2-n(n+l)/2 free elements. 1 7 4 . In order to be just-identified, the number of restrictions has to be 12 in this A and B matrix. But 13 restriction was given and it is over-identified. R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 133 ' 1 0 0" X 0 0 " A = 0 1 0 , B = 0 b2 2 0 K .a n a 32 K 0 ^33 j Based on this structural constraint, we will calculate the two Forecast Error Variance Decompositions (FEVD) that are Cholesky FEVD and Structural FEVD. Then we will compare the Cholesky FEVD and Structural FEVD to see how the interrelationships among variables influence each other over the structural break in the mid-1980s in two different structural impositions. (b) Comparing Forecast Error Variance Decompositions Let's compare the FEVD in two different orderings in two different periods for orthogonalized innovations of Cholesky decomposition and structural innovations. (1) FEVD based on lower triangular Cholesky decomposition The following four tables 3-17 to 3-20 present results of the forecast error variance decomposition with orthogonalized innovations of lower triangular Cholesky decomposition. There are two different sample periods divided over the mid 1980s. As we presented the structural changes in each equation in the previous sections, it is desirable to compare how the variations of each variable in response to the innovations are different over the two structurally different periods. We also set different orderings to probe whether the forecast error variance decomposition (FEVD) values are affected and how different it is. At table 3-17, the left-column orderings are export first, loan and the spread of the loan in the last. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 134 Right-column orderings are the ESFP loan first, exports, and the spread of the loan in the last. As we mentioned in the "Ordering of the Variables" section, it is to make sense of putting the spread of the loan in the last. The economic model, here, argues that export commitment would be affected by the ESFP loan rewarding system, and the ESFP loan is also influenced by the exports commitments. Therefore, putting the exports and the loans alternatively is a priori. <Table 3-17> First sample period: 1967.1-1984.4 Orderings Orderings Vars Lead EX Loan SOR Vars Loan EX SOR EX 1 1.00 0 0 Loan 1.00 0 0 2 0.98 0.052 0.001 0.89 0.000 0.100 3 0.91 0.067 0.261 0.86 0.003 0.138 4 0.80 0.112 0.091 0.74 0.061 0.203 5 0.81 0.128 0.066 0.73 0.066 0.207 Loan 1 0.26 0.764 0 DEX 0.24 0.764 0 2 0.21 0.687 0.100 0.21 0.787 0.001 3 0.20 0.660 0.138 0.28 0.697 0.269 4 0.26 0.533 0.203 0.32 0.592 0.091 5 0.27 0.525 0.207 0.21 0.720 0.066 SOR 1 0.05 0.002 0.942 SOR 0.02 0.035 0.942 2 0.12 0.003 0.878 0.03 0.096 0.878 3 0.14 0.003 0.861 0.03 0.114 0.861 4 0.12 0.002 0.873 0.02 0.103 0.873 5 0.13 0.002 0.869 0.03 0.103 0.870 1. Usually the leads are 25 percent of t le number of samp es, here, just five leads are presented for the purpose of the table space and 1 year of 4 quarters. 2. The data: export and export loan are the first differenced of level variables. Let's compare the FEDV proportions in the first and second sample periods in the left-orderings. At the table 3-17 and table 3-18, the left-orderings assume that it is more likely that an innovation in the export will lead to an innovation in the ESFP loans contemporaneously than that an innovation in the ESFP loans will lead to a R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 135 contemporaneous innovation in the exports commitment. Under such priori assumption, at the first sample period, the innovation of the ESFP loan is more explained by the export commitment than in the second period. For example, at the first lead, an innovation in the ESFP loan is explained by about 26% of exports, but in the second sample period, it is just explained by 11 % of the exports. Throughout the five leads, this type of different FEDV is presented in the same way. Therefore, we would conclude that after the structural change, the variations in the ESFP loan is not explained as much as it was explained by the exports. This suggests the fact that the government considers more other factors such as the international pressures, political agenda, and the large enterprises or SMEs than the export commitment. <Table 3-18> Second sample period : 1985.1-1996.4 Orderings Orderings Vars lead EX Loan SOR Vars Loan EX SOR EX 1 1.00 0 0 Loan 1.00 0 0 2 0.93 0.070 0.000 0.98 0.008 0.009 3 0.92 0.072 0.010 0.96 0.008 0.031 4 0.87 0.125 0.012 0.95 0.008 0.039 5 0.88 0.124 0.016 0.94 0.016 0.048 Loan 1 0.11 0.888 0 EX 0.112 0.888 0 2 0.15 0.837 0.009 0.21 0.793 0.000 3 0.15 0.820 0.031 0.20 0.789 0.010 4 0.15 0.812 0.039 0.21 0.784 0.010 5 0.15 0.795 0.048 0.18 0.805 0.020 SOR 1 0.00 0.038 0.954 SOR 0.02 0.022 0.953 2 0.00 0.048 0.947 0.03 0.019 0.947 3 0.01 0.049 0.934 0.05 0.020 0.934 4 0.05 0.058 0.895 0.07 0.035 0.895 5 0.07 0.056 0.866 0.07 0.064 0.866 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 136 Let's compare the FEDV proportions in the first and second sample periods in the right-column orderings. At the table 3-17 and table 3-18, the right-orderings assume that it is more likely that an innovation in the ESFP loans will lead to an innovation in the export commitment contemporaneously than that an innovation in the export will lead to a contemporaneous innovation in the ESFP loans. Under such priori assumption, at the first sample period, the innovation of the export explained by the ESFP loans more than the second period. For example, at the first lead, an innovation in the exports is explained by about 23% of ESFP loans, but in the second sample period, it is just explained by 11 % of the ESFP loans. Throughout the five leads, an innovation in the exports is explained averagely by 25 percent in the first sample period (before the structural change) and is explained averagely by 18 percent in the second sample period (after the structural change). The variations of the exports are less explained by the loans after the structural change in the ESFP loan system. Therefore, we may conclude there is a change in the firm's response to the ESFP loan system over the structural breaks since mid- 1980s. Firms maybe consider more other factors such as foreign investment, speculative financial investments, property purchases or R&D than the ESFP loans. The following two tables 3-19 and 3-20 are based on the VAR model with exogenous variables1 7 5 . The two exogenous variables are the export price index and 175. A VAR process can be affected by other observable variables that are determined outside the system of interest. Such variables are called exogenous (independent) variables. Exogenous variable can be stochastic or nonstochastic. The process can also be affected by the lags of exogenous R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 137 the wage index. Considering these two variables as exogenous variables in the VAR makes sense in that the Korean economy is small open economy. <Table 3-19> Two exogenous variables of EPI, DWI: First period sample: 1967.1-1984.4 Orderings Ore erings Vars lead EX Loan SOR Vars Loan EX SOR EX 1 1.00 0 0 Loan 1.00 0 0 2 0.99 0.003 0.002 0.85 0.011 0.141 3 0.96 0.013 0.025 0.81 0.011 0.183 4 0.88 0.071 0.053 0.72 0.081 0.201 5 0.85 0.101 0.053 0.72 0.080 0.203 Loan 1 0.21 0.792 0 EX 0.21 0.792 0 2 0.19 0.668 0.141 0.22 0.774 0.002 3 0.17 0.642 0.183 0.24 0.732 0.025 4 0.25 0.546 0.201 0.30 0.646 0.053 5 0.25 0.545 0.202 0.23 0.715 0.053 SOR 1 0.03 0.002 0.967 SOR 0.01 0.026 0.968 2 0.07 0.051 0.883 0.01 0.102 0.883 3 0.09 0.096 0.818 0.03 0.155 0.818 4 0.09 0.092 0.815 0.02 0.161 0.815 5 0.09 0.088 0.821 0.02 0.156 0.821 When we compare the FEDV proportions in the left-column1 7 6 orderings of two tables 3-19 and 3-20, an innovation in the ESFP loan after the structural change is variables. Here we do not give any lags to the exogenous variables. South Korea is small open economy. It is appropriate to assume Export Price Index and the Wage index are exogenous variables. The model is called as “VARX”., The VARX(p,s) model is written as P S y, =<?+£*/ yt-i+ T jx> -> +£< f=0 where y, = (yu,y 2t,y 3 ,)'and x, = (xl(,x2 ,)' SAS v8.2, in the Varmax procedure, Spliid, H.(1983), “A fast estimation for the Vector Autoregressive Moving Average Models with exogenous variables,” Journal of the American Statistical Association, 78, pp.843-849 1 7 6 . That is to say, the orderings are DEX, DLO, and SOR. The assumption of the orderings is the same as the previous two tables 3-14 and 3-15. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 138 found to be less explained than before the structural break by an innovation in the export commitments. Before the structural change, over the five leads, the average proportion of the loan FEDV is about 22 percent contemporaneously explained by the exports, but after the structural change, an innovation in the export explains just about 15 percent for an innovation in the loans. <Table 3-20> Two exogenous variables of EPI, DWI: Second period sample: 1985.1- 1996.4 Ordering s Orderings Vars lead EX LOAN SOR Vars LOAN EX SOR EX 1 1.00 0 0 LOAN 1.00 0 0 2 0.95 0.050 0.000 0.981 0.012 0.007 3 0.94 0.050 0.008 0.96 0.017 0.024 4 0.90 0.088 0.009 0.94 0.023 0.032 5 0.90 0.073 0.012 0.91 0.057 0.037 LOAN 1 0.09 0.907 0 EXPO RT 0.09 0.907 0 2 0.14 0.850 0.007 0.16 0.839 0.000 3 0.16 0.821 0.024 0.15 0.841 0.008 4 0.16 0.810 0.032 0.14 0.843 0.008 5 0.19 0.770 0.037 0.16 0.853 0.011 SOR 1 0.01 0.048 0.945 SOR 0.03 0.021 0.945 2 0.00 0.059 0.936 0.04 0.019 0.936 3 0.02 0.064 0.920 0.06 0.019 0.921 4 0.04 0.080 0.883 0.09 0.028 0.883 5 0.06 0.078 0.859 0.09 0.044 0.866 In the right-column orderings1 7 7 , comparing the explained innovations of the exports by the ESFP loans in the above tables, the ESFP loans contemporaneously explain the innovation in the export with 21 % at the first lead, 22 % at the second lead and 30 % at the third lead in the table 3-19. But in the table 3-20, an innovation 177. It means that, DLO, DEX and SOR. The assumption of the orderings is the same as R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 139 in the ESFP loans explains an innovation in the export with 9% at the first lead, 16% at the second lead and 15% at the third lead in the table 3-20. Because of the incentive structure changes, the variations of the firms' export commitment are less influenced by the ESFP loans in the second sample period. (2) Structural FEVD: On Exports and Export Support Loans (i) Effect of an innovation to Export Support Loan on Exports: Compare SFEVD of Before the Structural Adjustment and After the Structural Adjustment In the table 3-21 and 3-22, we see how innovations to export support loan affect differently the innovation in exports over the structural adjustments. <Table 3-21> Structural FEVD before the break: Orderings: Loan, Exports, SOR Response Variable SFEVD from Impulses Export Step Loan Export SOR 1 0 1 0 2 0.149169 0.850076 0.000755 3 0.186909 0.796603 0.016488 4 0.201091 0.782442 0.016467 5 0.20317 0.777608 0.019222 Comparing each SFEVD before and after the adjustment, like we confirmed the smaller forecast error variance in the Cholesky FEVD, we see the smaller structural innovations to export support loan on exports after the break. In table 3-21, at the fourth step the SFEVD of loan innovations on export is about 20%. But in the table 3-22, the SFEVD of loan innovations on export is about 5% which is by far smaller. the previous tables 3-17, 3-18 and the section “Ordering of the Variables”. R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 140 The influence of loan is much smaller after the structural adjustment in the export. This result confirms the expectation of the model which is that after the structural adjustment in the incentive structure, a representative firm will be less influenced by the policy loan. <Table 3-22> Structural FEVD after the break: Orderings: Loan, Exports, SOR Response Variable SFEVD from Impulses Export Step Loan Export SOR 1 0. 1 0 2 0.070682 0.861744 0.067575 3 0.065308 0.861453 0.0723239 4 0.057872 0.870934 0.071194 5 0.06357 0.861275 0.075155 (ii) Effect of an innovation to Exports on Export Support loan: Compare SFEVD of Before the Structural Adjustment and After the Structural Adjustment At this section, we investigate how innovations to export affect differently the innovation in exports over the structural adjustment period. We compare the SFEVD of loan from each impulse in the orderings of export, loan, and SOR. In table 3-23, the SFEVD of loan at the fourth step is explained by about 17% of export. But in table 3-24, the forecast error of loans is explained by 1% of innovations of export. <Table 3-23> Structural FEVD before the break, Orderings: Export, Loan, SOR Response Variable SFEVD from Impulses Loan Step Export Loan SOR R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 141 1 0 1 0 2 0.074373 0.88825 0.037377 3 0.165492 0.789921 0.044586 4 0.166826 0.781533 0.51642 5 0.169622 0.778658 0.051719 Table 3-23 continued. After the structural adjustment, the variation of export support loan was explained much less by the variation of export. This result confirms the expectation of this model. After the structural break in the mid-1980s, the variations in the ESFP loan is not explained as much as it was explained by the exports. <Table 3-24> Structural FEVD after the break, Orderings: Export, Loan, SOR Response Variable SFEVD from Impulses Loan Step Export Loan SOR 1 0 1 0 2 0.001602 0.995664 0.002734 3 0.009957 0.986845 0.003198 4 0.010368 0.986285 0.003347 5 0.01116 0.981117 0.007723 This result suggests the fact that the government considers more other factors such as the international pressures than the export commitment when the government makes a policy loan decision. IV. The Conclusion My puzzle started with the question of what is the superior institution for growth. I argue that the institution should have the high incentive system which gives consistently high motivation for the firms to commit full efforts to exports. In R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 142 small open developing countries, one of the economic engines is the exports. Exports lead investment, economic growth and efficiency. Once there is a slower growth in exports, this will slow economic growth. I assume this model is operating in the small open developing economy and exports matter in this kind of economy. Then I assume this export commitment by firms, which have been used to moral hazard, will be influenced by the incentive system given by the government which has been taking care of many different agendas, international pressures, and its own peoples' wealth. This economic model expects if the government gives less incentive system to the firms which is used to moral hazard and government leaderships in economic decisions, then this firm will commit less efforts to export commitments, even though the new government policy - less incentive system - is on the way to the economic reforms which are price reforms so called "Economic Liberalizations". In the short-run, unwanted economic outcomes such as economic crisis may happen due to less export commitment. Firm has a motivation, instead of increasing the export commitment, to increase deep pockets of speculative financial assets or other investment opportunities such as real estates investments, when firms maximize objective function under different incentive system given by the government. The questions I am raising here are how a firm and government react to each other under heterogeneous interests and moral hazard situation, and how their decisions are related to financial reform. Does the reform always produce R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 143 economic growth of an economy which gets used to moral hazard situation? My argument is that, even if the countries followed the financial market reforms well, under the moral hazard situations to which the firms have become accustomed to, when the reform lacks a growth-promoting quality1 7 8 , the economy may experience an unwanted crisis. I set up the government reward structure with the moral hazard model and introduced the idea of the Stackelberg leadership model under the game theoretic structure. The firms and the government are assumed to play this game over two different periods of pre and post-structural adjustment. In this theoretical model, I showed that when the government changes the incentive parameter, a firm would respond to this incentive parameter (growth- promoting quality) with higher or lower commitments of the exports. In a political economy framework, the government's interests and the firms' interests are different and there would be a possibility of conflicting interests over different periods. When there were international pressures on the government export support policy, the model shows that the government decreases the incentive parameter and this negatively affects the firms' commitment to exports. This provides the explanations how the incentive schemes influence firms' unobserved commitments in the theoretical model. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. In descriptive political economy chapter, we review how the export support financial policy has been changed under international political economy factors and how this different ESFP loan systems influence the export commitment over the pre- and post-structural adjustment in four different time periods. We investigated the policy changes and related economic data which were used to do empirical analysis in chapter III. In this chapter, we see the changes in international political economy horizons and how these changes have influences on the government ESFP loans over two different pre- and post-adjustment periods. We summarize the ESFP characteristics over the periods. The objective of the chapter III is to investigate the inter-relationships between the firms' export commitment and the government's incentive structure and thereby to suggest that the political economy1 7 9 matters for economic reforms and economic growth. Therefore, we expect the Structural Change Test, ANOVA (Analysis of Variance), Polynomial Distributed Lag Model, a VAR (Vector Autoregression) and Structural VAR to support the argument that the lowered incentive schemes would lower the firm's commitments to exports and that it might lead to the crisis. In the empirical analysis, we started the analysis with testing the stationarity of the times series data. And using this differenced data, we showed there were structural changes in the ESFP in the middle of 1980s with the structural test of 178. Which I call incentive parameter. 179. Heterogeneous Interests in the objective functions for the economic decision makers. perm ission of the copyright owner. Further reproduction prohibited without perm ission. 145 dummy variables. This change lowered the incentive parameter to lower the firms' commitment to exports. This lowered incentive parameter finally lowered exports growth, too. It is proved that firms respond to the change in the incentive parameter with the analysis of variance. All of the structural interaction dummy variables negatively affected the incentive parameter and the firm's commitment. In the polynomial distributed lag models, we showed the lagged exports significantly influenced the export support loans within six months and the lagged loans significantly influenced exports within six months, too. In addition to the results in the single equation methods, we employed the Hausman specification test in a multi-equation system to test the endogeniety among variables. Due to the result of endogeneity, the VAR model was introduced to investigate the inter-relationships among the exports, ESFP loans, the interest rates' spread and other variables. The benefit of introducing this standard VAR model is to treat all variables endogenously so that we can get away from the criticism of so called "Incredible Identification Restrictions". In a VAR(p), we found loans and exports significantly influenced each other with some lags. Foreign pressures also affect the loan decisions and have significances on the equations. The effect of export price index was also pronounced. Granger causality test showed ESFP loans Granger-caused exports commitments and exports Granger-caused the ESFP loans. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 146 With the Cholesky factorization in error terms, we introduced the Forecast Error Variance Decomposition (FEDV) to investigate the changes in variations in an innovation of each variable. This result shows that the decreased proportions in an innovation of exports and in an innovation of ESFP loans which are explained by each other after the structural adjustment in the mid-1980s. This reinforces the argument that the firms and the government have been reacting to each other through the incentive structures under the moral hazard environments over pre- and post-structural adjustment periods. The lowered incentive scheme lowered the firm's efforts to export commitments and this lowered the economic growth. Then, I introduced Structural VAR model which has distinctive benefits of reflecting the theoretical assumptions of the suggested economic model. A standard VAR model uses the triangular restrictions on the error terms which may be too specific on a specific model. For this economic model presented here, ESFP loans sequentially affect the firms' exports commitments and exports commitment sequentially affects the ESFP loans. The interest rates simultaneously are determined by the market. We use these structural assumptions to impose the restrictions in the parameters of the structural VAR model. Fortunately, the structural VAR results show very similar to the Cholesky decomposition errors in a standard VAR. It confirms the argument that the different incentive schemes produce different firms' export commitments. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 147 And it suggests, under moral hazard model, once the government's strong incentive reward system was removed, in order to keep the continuous economic growth, the economy may need to provide different kind of rewarding incentive structures for the firms who maximizes objective functions after the structural adjustment. Just only the price reform without enhanced incentive schemes may not be the all-mighty solutions to the economic growth. If the economy is not given a proper growth promoting incentive, the economy would observe less commitment to exports and this may lead to the unwanted macroeconomic performances such as crisis. Therefore, there would be a suggestion that the political economy of heterogeneous objective functions with different policy incentive structure under moral hazard could play an important role in determining the different directions of the economic outcomes in the stages of economic reforms by the theoretical and empirical analysis. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 148 References Abdullah, Dewan, 1998, "Money Growth Variability and Stock Returns: An Innovations Accounting Analysis", International Economic Journal, Voll2, 89-102 Ahmed, Shaghil and Murthy, Radha, 1994, "Money, Output, and Real Business Cycles in a Small Open Economy", The Canadian Journal of Economics, Vol. 27, No.4, 982-993 Amisano, Gianni and Giannini, Carlo, Topics in Structural VAR econometrics, 1997, Springer-Verlag, chapter 1-6 Balke and Gordon, R.J, 1986, Historical Data, in The American Business Cycle, ed. The University of Chicago Press Bergeijk, Sinderen, and Volland, 1999. Structural Reform in open economies. Bekaert, Geert, and Harvey, Campbell, 2003, "Emerging Markets Finance", Journal of Empirical Finance, Vol 10, 3-55 Berkowitz, Daniel and Dejong, David, 2003, "Policy reform and growth in post- Soviet Russia", European Economic Review, vol 47,337-352 Bianchi, Marco, 1995, "Granger Causality in the presence of structural changes", WP33, Bank of England. Blanchard, Oliver, and Quah, Danny, 1989, "The Dynamic Effects of Aggregate Demand and Supply Disturbances", The American Economic Review, Vol.79, No.4, 655-673 Blanchard, Olivier, 1989,"A Traditional Interpretation of Macroeconomic Fluctuations", The American Economic Review, vol 79, N o.5,1146-1164 Blangiewicz, Maria and Charemza, Wojciech, 1999, "East European Economic Reform som e sim ulations on a structural VAR m odel", Journal o f Policy M odeling, vol 23,147-160 Brander and Spencer, International R&D Rivalry and Industrial Strategy, 1983, 703-722 Buckle, Kim, Kirkham, McLellan and Sharma, 2002, "A Structural VAR model of the New Zealand business cycle", New Zealand Treasury Working Paper 0/26 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 149 Burbidge and Harrison, 1985, "Innovation Accounting for the Impact of Fluctuations in U.S. variables on the Canadian Economy", The Canadian Journal of Economics, Vol. 18, No.4,784-798 Canova, Fabio, 1991, "The Sources of Financial Crisis: Pre-and Post-Fed Evidence", 1991, International Economic Review, Vol. 32, No.3, 689-713 Cochrane, John, 1998, "What do the VARs mean? Measuring the output effects of monetary policy", Journal of Monetary Economics, vol 41, 277-300 Cochrane, John, 1994, "Permanent and Transitory Components of GNP and Stock Prices", The Quarterly Journal of Economics, Vol. 109, 241-265 Drazen, Allan, 2000. The Political Economy in Macroeconomics. P.l-30. Demetriades and Luintel, 2001, "Financial constraints in the South Korean Miracle," Journal of Development Economics 64, 459-479 Demetriades and Hussein, 1996, "Does financial development cause economic growth? Time-series evidence from 16 countries", Vol 51, Journal of Development Economics Demetriades and Luintel, 1996, "Financial development, economic growth and banking sector controls: Evidence from India", voll06, The Economic Journal, 359-374 Dewan, Abdullah, 1998, "Money Growth Variability and Stock Returns: An Innovations Accounting Analysis", International Economic Journal,Vol 12, 89-104 Enders, Walter, 2004, Applied Econometric Time Series, 2n d Ed, 239-318 Enders and Sandler, 1993, "The Effectiveness of Antiterrorism Policies: A Vector- autoregression Intervention Analysis", The American Political Science Review, Vol. 87, No. 4, 829-844 Evans, Charles and Kuttner, Kenneth, 1998, "Can VARs Describe Monetary Policy?", FRB of Chicago Working Paper Series Green, William, 2002, Econometric Analysis, p.250-282, p.558-662, fifth edition R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 150 Fajnzylber, Pablo, and Lederman, Daniel, 1999, "Economic Reforms and Total Factor Productivity Growth in Latin America and the Carribbean, 1950-95: An Empirical Note", Working Paper Fare, Rolf, Grosskopf, Shawna and Margaritis, Dimitri, 2002, "Economic reform and productivity growth: the case of Australia and New Zealand", Mathematics and computers in Simulation 59,143-152 Fernandez, Raquel and Rodrik, Dani, 1991, "Resistance to Reform: Status Quo Bias in the Presence of Individual-Specific Uncertainty", The American Economic Review, Vol.81,1146-1155 Friedman, Benjamin and Kuttner, Kenneth, 1992, "Money, Income, Prices, and Interest Rates", The American Economic Review, Vol 82, 472-492 Gelos, R and Werner, Alejandro, "Financial liberalization, credit constraints, and collateral: investment in the Mexican manufacturing sector", vo!67, Journal of Development Economics Giugale, Marcelo, and Korobow, Adam, 1999, "Shock persistence and the Choice of Foreign Exchange Regime: An empirical Note from Mexico", 3-16 Gottschalk, Jan, 2001, Working Paper no. 1072, "An Introduction into the SVAR Methodology: Indentification, Interpretation, and Limitations of SVAR models", 1- 40 Greene, Econometric Analysis, 2003, chapter 15,19, and 20. Griffiths, Hill and Judge, 1993, Learning and Practicing Econometrics, p.680-707 Gujarati, Damodari, 4th ed, 2003, Basic Econometrics, 441-867 Hamilton, James, 1994, Time Series Analysis, 318-336 Holmes, James and Hutton, Patricia, 1992, "A New Test of Money-Income causality", vol24, issue3, p338-355, Journal of Money, Credit and Banking. IMF Country Report No.03/81, March 2003, "Republic of Korea: Financial System Stability Assessment, including Reports on the Observance of Standards and Codes on the following topics: Monetary and Financial Policy Transparency, Banking Supervision, Securities Regulation, Insurance Regulation, Corporate Governance, and Payment Systems" R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 151 Jwa, Sung Hee, 2002, The Evolution of Large Corporations in Korea, 17-56,. 89-153 Jean-Jacques Laffont and David Martimort, The Theory of Incentives,2002, chapter 1, 2, 3,4,5 Johnston and Dinardo, 4th ed, 1997, Econometric Methods, 287-326 Kasa, Ken and Popper Helen, 1996, "Monetary Policy in Japan: A Structural VAR Analysis", Journal of the Japanese and International Economies 11, 275-295 Keating, John W., 1992, "Structural Approaches to Vector Autoregressions" KIEP, 1997, "The political Economy of Korea-U.S. Trade Conflict: Development and Challenges" Konya, Laszlo, Dec 2000, "Export-led growth or growth-driven export? New Evidence from Granger Causality Analysis on OECD countries", WP15, School of Applied Economics Laffont, Jean-Jacques, 2000, Incentives and Political Economy, Oxford University Press, 1-150 Li, Wei. 1997. "The Impact of economic Reform on the performance of Chinese state enterprises", The Journal of Political Economy. 1080-1106 Laffont, Jean-Jacques, and Martimort, David, 2002, The Theory of Incentives, Princeton University Press, chapter 1-5. Lin, Wen-Ling, 1995, "Japan's Financial Deregulation and Linkage of the Gensaki and Euroyen Deposit Markets", Journal of Applied Econometrics, Vol. 10, 447-467 Lutkepohl, Helmut, Introduction to Multiple Time Series Analysis, 2n d Edition, 1993, 9-116 Manchester, Joyce, 1989, "How Money Affects Real Output", Journal of Money, Credit and Banking, Vol.21, No.l, 16-32 Pindyck and Rubinfeld, 1991, Econometric Models and Economic Forecasts, 367-471 Rodrik, Dani, 1999. Institutions for high quality growth: what they are and how to acquire them. Working paper. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 152 Rodrik, Dani, 1997, Democracy and Economic performance, Conference paper. Rodrik, Dani, 1996, "Understanding Economic Policy Reform", Journal of Economic Literature, Vol 34,9-41 Runkel, David E., 1987, Staff Report 107, "Vector Autoregressions and Reality", FRB of Minneapolis Research Department. 1-13 Sachsida, Adolfo, 1999, A Causality Test Between Financial Deepening and Growth", SSRN Electronic Paper Collection Silberberg, Eugene, and Suen, Wing, 2001, The Structure of Economics, chapter 13-16 Sims, Goldfeld and Sachs, 1982, "Policy Analysis with Econometric Models", Brookings Paper on Economic Activity, Vol. 1982, No.l, 107-164 Sims, Christopher, 1980, "Macroeconomics and Reality", Econometrica, Vol.48,1-48 Sims, Christopher, Goldfeld, Stephen, and Sachs, Jeffrey, 1982, "Policy Analysis with Econometric Models", Brookings Papers on Economic Activity, Vol. 1982,107-164 Sims, Christopher, 1986, "Are Forecasting Models Usable for Policy Analysis?", FRB of Minneapolis Quarterly Review Spencer, David, 1989, "Does Money Matter? The Robustness of Evidence from Vector Autoregressions", Journal of Money, Credit, and Banking, Vol, 21, No.4, 442-454 Swanson, and Granger, 1997, "Impulse Response Function Based on a Causal Approach to Residual Orthogonalization in Vector Autoregressions", 357-367, Journal of the American Statistical Association, Vol, No.437 Sims, Christopher, 1980, "Macroeconomics and Reality", Econometrica, Vol48, No.l, 1-48 Sims, Christopher, 2002, "Structural VARs", Time Series Econometrics, Econ 513,1- 10 Sturgess, Brian and Wilson, Nicholas, 1984, "Advertising Expenditure and Aggregate Consumption in Britain and West Germany: An analysis of Causality", Managerial and Decision Economics, Vol 5, 219-227 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 153 Stock, James and Watson, Mark, 2001, "Vector Autoregressions", Journal of Economic Perspectives, vol 15,101-115 Taylor and Spriggs, 1989, "Effects of the Monetary Macro-Economy on Canadian Agricultural Prices", The Canadian Journal of Economics, Vol.22, 278-289 Williamson, John, 1994. The Political Economy of Policy Reform, pl-72 Bank of Korea, Statistical Year Book, each year KIET, The Theory and Practice of Economic Reform, 1999, Research Report 99-04 IMF, International Financial Statistics, each year R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The fortunes of fishes and fishers: The political economy of innovation in Atlantic resource management

PDF

Political economy of financial liberalization in emerging markets: A comparative study of South Korea and Taiwan in the 1990s

PDF

The political economy of Turkish Islamism and secularism: The role of preference falsification in the establishment of Kemalist secularism and the polarization of Turkish public opinion on the pu...

PDF

The gravity model at work: An empirical study on the effects of governance on bilateral trade flows with special emphasis on intra-Arab trade

PDF

The corporate economy and growth of corporations in Taiwan

PDF

The transformation of Taiwanese political economy since the mid -1980's: From authoritarian to quasi -democratic state

PDF

The political economy of procyclical fiscal policy in Mexico, 1970--1988

PDF

The outcomes of trade conflicts and negotiations between the United States and South Korea, 1980--2000

PDF

Restructuring and performance in India's electricity sector

PDF

Three essays on democracy and economic policy

PDF

Carrots or sticks: The Taiwanese government's contrasting policy approaches to banking sector liberalization

PDF

Analysis of the role of government in Taiwan's industrialization and economic development

PDF

The determinants of child growth retardation in India

PDF

Before and after reform: The effects of structural adjustment on women's employment in Latin America

PDF

The Japanese main bank system: A transaction cost approach

PDF

A comparative examination of the impact of business -government relations on labor market reform in Egypt and Mexico, 1975--1995

PDF

Trends in mortality in Turkey, 1960--1995

PDF

Three essays on asymmetric information and public policy: Information, the environment, and the provision of defense goods

PDF

The implementation of endangered species policy: An analysis of the 'takings' issue in specific contexts

PDF

Economic sanctions, domestic audiences, and international conflict

Asset Metadata

Creator Lee, Junkyu (author)

Core Title Political economy of economic reform, quality of reform and economic performance: Reform matters, but consistent incentives matter more

Degree Doctor of Philosophy

Degree Program Political Economy and Public Policy

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Economics, General,OAI-PMH Harvest,political science, general,political science, international law and relations

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-413392

Unique identifier UC11335830

Identifier 3145226.pdf (filename),usctheses-c16-413392 (legacy record id)

Legacy Identifier 3145226.pdf

Dmrecord 413392

Document Type Dissertation

Rights Lee, Junkyu

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA