External shocks, adjustment policies, and investment : illustrations from a forward-looking CGE model of the Philippines

_ _ _~~~~ ~~ ~~ ~ /. l W ( )g / 9 J) | Policy, Research, and Exiernal Affairs WORKING PAPERS Public Economics Country Economics Department The World Bank August 1991 WPS 737 External Shocks, Adjustment Policies, and Investment Illustrations from a Forward-looking OGE Model of the Philippines Delfin S. Go The rapid increase in investment and external debt of middle- income countries like the Philippines during the 1970s was perfectly "rational" behavior, given existing policies. However, these countries could have done better with an appropriate Imlix of adjustment policies. The paper highlights the intertemporal tradeoffs of tariff refonm, emphasizing the need for complemen- tary measures to ease macro imbalances and short-term disloca- tions of the protected sector. ThePohcy,Research.and ExtemalAffairs Complcx distnhutes l'RI. Working Papers todisserunateLhefindings of %ork in progress and to encourage the ecxhange of ideas among Bank staff and all others interested in development issues. T'hese papers carry thC names of the authors, reflect only their viows, and should be used and cited accordingly. The findings. iiterprctations. and conclusions are the authors' own. They should not be attributed to the World Bank. its Board of Directors, its managemein, or any of iLs member countnes.  I Policy, Research, and External Affairs I Public Economics WPS 737 T; ,i' 1'. Ipe r a product ol' thc Public Economics Division, Country Economiiics Dcpartment -- is part of a aice ..lcort in PRE to cxaminc open-cconomy tax reforin and adjusiment policies in dcveloping ,'W tlir s. Copies are available free from thc World Bank, 1818 II Sircc NW, Washington DC 20433. 1 1 contact Ann l3halla, room N 10-059, exteinSioin 37699 (48 pages). k t,dcx cl)pcd a m;odel that integrates from the perspective of the interest rate shock of' tem 1 I iuporal an:1d forward-looking behavior in 1979. The expectation that debts could be i;, -, imcni ii anid conisumption decisions in a stacked indcfini:cly was clearly wrong and the , lliEcc ioat general equilibrium framework initcrest rate shock after 1979 came as a surpris- . iabkh al t) developing counitries. It formulates to many. .;IJ IP""" u ll inminite horizon growth mo(lel to .Ofi,,}t. tilt adjustment, growth, and debt The results demonstrate both the prrmise . ! hmi. of' a i-iddle-incomic couiltry, which Go and danger ol liberalizationi, an adjustment ,:l,<' (I .1mg data lol iltc Plhilippincs. policy ofrten recomiimenided in the 1980s. Tariff' reform shows ihc cxpected benefits in the i t . indalions illustrate the importance of' primary sector and to some degree in exports. .: xx aid looking behavior in conisump- But it may easily lead to a contractioni in the 1! n " nx cstInent. Tlhc) shiow how an import protccted sector like manufacturing, a decline in -ht k could lea(i to an investimnct boom tax revenues, more current account deficits, and i I i I C xpalsion of forcigin debt, as happened more debt. Other mcasures ar needed. , i i. i d s.c cnmtic, to many develophing couIn- * In general, Go 1iiids that a combiniationi of policies is effective in maintaininlg growth and 1! apid cxpansion is hard to explain exports wvithout a rapid accumulation of'debts- 'i!; umtoatino the dynamic competitive cven duiing a permanent import price shock. In : btween domestic an1d ioreign goods fact, an import price slhock is an attractive .i!n in itkeponisC to this shock. TlC increasc occasion lor tariff reform, as it cases prcssures j: !:. Ill account dcficits and ioreigni debt aitcr on domestic piices, prevents exports f'roimi I I't price shock, for examiiple, is certainly declining too much, and docs not lead to an : thinm s 'o(ild be implied by just thc in- expansion of import demand thal normally I inl lircltin prices. Go concludes, among accompanies a tariff reduction. Cotabined with other policies, tariff refonn vi ildion is a kcy factor. Contrary to Ole could rechannel investment and resources toward t i: ,.U I sugge:stionl that anl economy should the morc tradable sectors an(i exports cain be . ail' and colntrac in response to a pemianent emphalsized and increased. If domestic resources 1111m,i I pl)iic shock, tile behavior- suggested in a are also mobilized tirough increased tax collec- aJ. i, t x ith raitional cxpeciationis in investment tioin, thc combir'.cd effect wvill be to reduce or !'i),;l,, 1i' that thC opix)site can be rtrc. slow the accumulation of foreign debt. 1 -,].wcladtions prevailing in the imnid-sevenlties 1- J) 'hit ilic energy crisis was pennanent, the In other words, middle-inicomc coLintries likc i C ea1' kjlvap oil were over, and petrodollars the Plhilippines IImissed a goldcie opportunity for v. o ntilt iitit tuc to he available. But condiitions policy reform in the 1 970s anid found it harder to ;. IanTd. 'I'Thc actioins of thc high-debt implemnIt adjustimicni policies unide less j 'i, ng countlies apiear "inmpudent" viewed favor-able circumistances in the 1980s. i\ 1 LI iu per S.ries. dissemrnates the findings of work undeles \ay in tlie Bank's l'olicy, Researth, and Extemal ie l u .i Atl obJ c tie eol the scrie. cis to get these findinlgs out qulickly, evesn if presentaltionls arc less thlan fullly pxlished. ' ]. s1] F,, intctletitions. ,aid conclusionn in ilesc papers do not neces-arily represent official BRnk policy. Produced by thc l'RE D)issemination Ceniter  Contents 1 Introduction 1 1.1 General Framework .. . .. 2 1.2 Organizatiou of the Paper ....... . . . . . . . . . . . . . . . . . . . .. 6 2 Model 6 2.1 Consumption . .......... . . . . . ............... 6 2.2 Investment ........................ . ..... 8 2.3 The Intra-period CGE .i............................. 10 2.4 Equilibrium Conditions ..... . ..................... 11 2.5 Solution Strategy ............................... 12 3 External Shocks 14 3.1 The Impact of an Import Price Increase ....... ............. . 16 3.2 Features in the Dynamic Behavior ...... . . . 19 3.2.1 The Simultaneity of Investment and Saviz&gs .1.9... . . . . . . . . . 19 3.2.2 The Intertemporal behavior ...... . . . . . . . . . . . . . . . . . 20 3.2.3 The Forward-Looking behavior ...... . . . . . . . . . . . . . . . 21 4 Adjustment Policies 23 4.1 Tariff Reform .................................. . 24 4.2 Policies During an Import Price Shock ...... . . . . . . . . . . . . . . . 26 S Conclusions 28 Figures 32 References 38 Appendices 41 ListofEquations .......... 41 Glossary ......... . 46 The analysis of this paper is based on the author's Ph.D. dissrtation, which received an individual grant from Ford Foundation, Philippines. Special thanks re due to Shantaynan Devarajan for his valuable comments and to Hollis Chenery and Dwight Perkins, the other membewr of the Ph.D. Committee. This paper also benefited from various comments in the Bank, among others, from Martha de Mdo, Norma Hicks, Emmanuel Jimenez, Javad Khali[zadeh-Shirasi, Johannes Linn, Pradeep Mitra, and Zmarak ShalizL  1 Introduction The economic difficulties triggered by high oil prices during the brief Persian Gulf crisis underline the continuing importance of external developments and adjustment policies in shaping the economic performances of developing countries. It may be recalled that during the seventies, a number of middle-income and semi-industrial countries actually did well after the 1974 oil crisis but had difficulties after the 1979 oil and interest rate shocks. Hindsight tells us that many of these developing economies over-invested and over-borrowed in the mid-seventies but had to contract when faced with constraints in further borrowing after 1979.1 The implementation of adjustment policies in the eighties has led to the problem of slower growth, characterized by a drop of investment as a ratio to GDP.2 A key factor in these experiences is the behavior of investment and its sensitivity to price changes brought about by external shocks and policy changes. Why did investment in some middle-income countries rose so much in the seventies only to have severe debt problems later? Why did investment decline when adjustment policies, such as trade reform, are adopted? These policies, aimed primarily at making investment and growth more efficient and less dependent on foreign resources in the eighties, are constrained by their adverse impact on macro balances and investment activities in the short term. This study attempts to explore this type of investment and growth responses in a middle-income country like the Philippines by using a dynamic general equilibrium analysis. 'These observations were made by several studies such as Balawsa (1986) and Balassa axid McCarthy (1984). 2See, for example, reviews of adjustment lending experiences of the World Bank which contain major concerns for the recovery of investment and its praductivity in Adjustment Lending - An Evaluation of Ten Years of Experience (1988) and Adjustment Lending Policies for Sustainale Groawth (1990). 1  1.1 General Framework This study is distinguished from other research on developing issues by the integration of an intertemporal and forward-looking behavior in investment and consumption decisiona in a computable general equilibrium (CGE) kramew3rk app!"table to developing countries. It formulates and uses an infinite-horizon growth model to examine the adjustment, growth and debt problems of a middle-income country using Philippine data for illustrations. This growth model is maltisectoral and intertemporal. It combines the intersectoral efficiency in the allocation of resources resulting from market dearing and endogenous prices of a general equilibrium model with the intertemporal efficiency generated from the dynamic optimization of the firm and the utility of an aggregate consumer. This quantitative framework allows us to explore several issues about the growth and investment behavior of a middle-income country in a way not possible in models without expectations. While there is growing number of CGE applications to developing issues, for example, Devarajan and Sierra's examination of Thailand (December 1986) and Lewis' study of Turkey (July 1986), these CGE models are static in nature and are ill-suited for analyzing the dynamic processes of growth. I In particular, the effects of external shocks and adjustment policies on investment and growth are inherently dynamic -- an import price shock or a tariff reform changes the expectation of profits permanently and hence the level and pattern of investment. In these previous CGE studies, an economy's movement in time is described by a series of static equilibria strung together by updates in the levels of factor endowments. These updates axe ad hoc and exogenous to the tatonnement. In relation aThe strengths and weaknesses of economy-wide models in analyzing development issues are discused in Beil and Suinivaaan (1984). 2  to the observed patterns of middle-income countries during the seventies, these previous CGE studies stop short of explaining why investments (hence debts) were high after the first oil shock in 1974 and low after the second shock in 1979. Investment levels were either exogenousty fixed at their high levels or d,iven low by arbitrarily setting them equal to a lower level of domestic and foreign savings as required by the macro-dosure of the model. One effect that these models were designed to measure - the intersectoral reallocation of resources caused by the market price shifts - was often swamped by exogenous changes in investments.4 To examine these dynamic issues, a forward-looking behavior in investment and con- sumption is embedded in a multisectoral and general equilibrium representation of a devel- oping economy. Unlike the static case, the solution of the model is a dynamic competitive path in the sense that it is an intertemporal path along which the economy is always in general equilibrium. It is characterized by three key features - the simultaneity of invest- ment and saving decisions, the intertemporal and forward-looking behavior, and a general equilibrium applicable to developing countries.5 First, savings and investment decisions are not only intertemporal but separate and simultaneous. Investment is neither fixed as in static CGE applications nor dynamically passive as in studies concerned mainly with optimal borrowing, for example, Kharas and Shishido's (1985) study of Thailand and Devarajan's (1986) examination of Korea. The 'While it is possible to set the various elasticities of substitution very low to have strong reallocation effects, convergence problems are often encountered first as the models become rigid and approach Leonitf's fixed-coefficient world. The effects of relative price shocks in a static CGE generally run their course in a single period and have less impact on real income over time than changes in factor supplies. 'There are some recent studies that utilize intertemporal models with forward-looking investment be- havior, but they are not applied to development problenm. For examples, Bruno and Sachs (1985) and Summer(1985) have forward looking investment incorporated in very macro simulation models with applica- tions to the OECD countries or the United States. Wilcoxen (1986) discusses forward looking investment in general equilibrium models and their numerical calculation using linearization. Goulder ani Summer (1989) incorporates this type of investment behavior iu a multisectoral general equilibrium model for the United Statez and examines tax incentive issues for investment. 3  solution is decentralized. At each point in time, consumption is an increasing function of wealth as in a choice-theoretic and life-cycle consumption model. Investment is an increasing function of Tobin's q, the ratio of the present value of profits to the cost of capital, and is subject to adjustment cost as in Hayashi (1982). The macro-closure and equality of savings and investments are brought about by an endogenous adjustment in the level of foreign borrowing, supplied at a given world interest rate. We assumed that a middle- income country have sufficient access to the world capital markets in the s3eventies. When foreign borrowing is constrained as in the eighties, we explore domestic sources of savings and examine the impact on debt accumulation. Second, the model is dynamic and forward-looking. Consumption smoothing is carried out by the consumer in anticipation o'i changes in consumer prices. Similarly, capital ac- cumulation is started by the firms in response to future prices of net output and the cost of capital. In this model, consumer and producer prices are determined by the prices of domestic and foreign goods that are imperfect substitutes. Hence, the dynamic paths of consumption and investment will depend on the changing competitive conditions between domestic and foreign goods. These conditions are reflected in the appropriate real exchange rates and their impact on the intertemporal rates of transformation for consumption and investment. Moreover, because of the adjustment cost, the new steady-state capital stocks are approached gradually over time. The speed of adjustment in the production sector is not instantaneous. Third, it is a multisectoral and general equilibrium model with imperfect substitution. In addition to the Walrasian paradigm of price endogeneity and market clearing, a variety of imperfect substitution in the trade and labor markets are incorporated. Thus in each 4  period, changes in relative prices will affect resource allocation in the tradition of general equilibrium models applied to develoning cc antries. Finally, a middle income country is chosen because the higher per capita income reflects a more diversified economy with a more developed manufacturing sector capable of inde- pendent investment decisions. The analysis uses the Philippines as a case in point. Like other oil-importing and middle-income developing countries, the Philippines was hit hard by the four-fold oil price increase in 1974 and the world-wide inflation and stagnation after. Yet from 1974 to 1979, the Philippine economy grew by 6.5% annually which was higher than its long term growth of 5.5%. Much of this growth was fihanced by external borrowing. When foreign loans became costly in the eighties and adjustment policies were undertaken, the economy went through a serious financial crisis that was eventually marked by a 10% decline in GNP and widespread unemployment during 1983-85. As in many such instances, the repercussions went beyond the economic sphere and contributed to a popular civilian- military rebellion that brought down a 20-year strongman government in early 1986. The size of the foreign debt in the Philippines was a staggering $26.2 billion in 1985. This was about 80% of its annual gross national product (GNP). Debt service required $1,257.0 mil- lion or about 27% of its annual merchandise exports. In recent years, moderate economic growth has been restored but the country is still faced with serious questions about its long- term policies and prospects. Taking the Philippines as an example and using a dynamic simulation model, can we explain its economic and investment performance? What could have been done differently? 5  1.2 Organization of the Paper The remaining sections of this paper are organized this way. Section 2 briefly describes the specification and implementation of the model. Section 3 examines the growth and invest- ment behavior of a middle-income coi'ntry in the seventies and early eighties by introducing an import price shock to the model. To examine whether the dynamic effect: generated by the proposed framework add new insights to the observed responses of the middle-income countries, we compare the results with those a static CGE model. Section 4 investigates some of the policy options while section 5 summarizes the findings and conclusions. 2 Model In this section, we briefly describe the framework with emphasis on the intertemporal con- sumption and investment.6 A list of equations and a glossary of terms are included in the appendix. A reference such as 'A20' means equation '20' in Appendix A. 2.1 Consumption The representative consumer maximizes his expected utility or the present value of the v tility of total consumption in each period which is expressed as follows: ma 1 C, -ve-Pt (1) This is a familiar homogeneous utility function which is additively separable and a case of constant elasticity of marginal utility v. This type of utility function has been used often by Ramsey, Frisch, Timbergen and others in the economic literature. Consumption eGo (1988) provides a more detailed discussion, including the derivation and calibration of various equa- tions in the model. 6  Is discounted by the consumer's rate of time preference p. In order to mantaiu dynamic consistency as posed by Strotz (1955-56), we assume that p is fixed through time. At each point in time, the consumption bundle C is defined by a Cobb-Douglas agegation of specific consumer goods C,. This is akin to an intra-period utility function with fixed expenditure shares a, (A2). The intertemporal condition for consumption determines the forward rate of growth of consumption in response to changes in the intertemporal rate rdg by which current consumption is transformed into future consumption7 - Ct rt-Pt (3) A large re makes future consumption cheaper and the rate of g-owth of consumption will increase, and vice-versa. The intertemporal rate rt is determined by the opportunity cost of savings, the cost of foreign borrowing (A22). This depends on the world interest rate i and the real exchange rate ec, by which import substitutes of domestic goods are traded to satisfy consumer demand. Aggregate consumption is function of wealth and can be derived from the optimal con- ditions and the wealth constraint. In the solution strategy however, we take advantage of certain conditions in which a central planning formulation will give identical results as a decentralized CGE with a single infinitely-lived consumer. Furthermore, the consumer bud- 7The Hamiltonian for this problem is H = e-1"[U(Ct) + 'y.(YC. + rctWt -PCtCt) (2) The optimality conditions from the maximum principle are as folows: a) OR-=