Southeast Asian Studies. VoL, 24. No. 4. Match 1987
A Medium-term Macroeconometric Model for Economic Planning in Indonesia.
Sei KURIBAYASHI**
a better consistency in the planned I Introduction macroeconomic aggregates. In this respect The five-year development plan which the present plan constitutes a step forward the Indonesian Government works out from the past practice." every five years is indicative by nature. The macroeconomic models consist of a setting policy guidelines for government macroeconometric model. an interindustry development policies. Three development model and two submodels. agriculture and plans have already been formulated and energy, as described in Fig. 1. The carried out since 1969, and implementation macroeconometric model comprises a core of the Fourth Five-year Development Plan model, a fiscal submodel. a monetary (Repelita IV) began in April, 1983. In submodel and a balance-of-payments Repelita IV, macroeconomic mociels were submodel. The core model and each employed in order to obtain consistency in submodel were constructed and tested macroeconomic aggregates in the economic independently at the first stage, then at framework of the plan. The plan .states: the later stage attempts.were made to "Repelita IV has attempted to employ integrate them. The monetary submodel, macroeconomic models as guidance for its however. was not connected with other broad quantitative estimates. This has models because of our shortage of com enabled the plan to take a better account puter capacity.I) Consequently, the macro- of the existing interdependencies and .econometric model. which consists of the interrelations among variables as well as core model, the fiscal submodel and the among sectors, with a view to obtaining balance-of-payments submodel, was used for the purpose of conducting simulations * I am grateful to Shinichi Ichimura, Kazumi for Repelita IV. The balance of payments Kobayashi, Sumimaru Odano, and Takao Oshika for many helpful conversations and submodel, however, was not completely useful comments. I also thank Adrianus Mooy and other staff members of A Quanti 1) The monetary submodel was successfully tative Study on the Medium/Long.term connected with the core model, but when Prospect of the Indonesian Economy for their we tried to integrate the core model with warm hospitality and help during my stay in the monetary submodel as well as the fiscal Indonesia. submodel, we encountered difficulties with ** ~*" t!:t. Economic Research Institute, Eco the computer capacity. For the monetary nomic Planning Agency, 3-1-1 Kasumigaseki submodel, see _Ezaki [1982] and Odano Chiyoda-ku, Tokyo 100, Japan ' [1983a].
350 S. KURIBAYASHI: A Medium-term Macroeconometric -Model for Economic Planning in Indonesia
nesian economy in the 1970s and Macroeconometric MOdel 1980s in terms of macroeconomic
Fiscal Submodel aggregates. Section 3 explains
the structure of the core model.2) Balance-of-payments Monetary Submodel Core Model In section 4 multiplier analysis Submodel is briefly presented.
Agriculture I Submodel n Economic Development in Indonesia Energy I Submodel. 1. Economic Development since Interindustry 1969 Model It goes without saying that a Fig. 1 Macroeconomic Model for Economic Planning necessary condition for model in Indonesia builders is to have a good grip integrated with the core model and the fiscal of the characteristics of economic develop submodel, but rather used for checking ment in the past and of problems and purposes, for it did not have feedback issues of the present and future. First, loop to influence the other models, and therefore, a brief description is given of exchange rate treated as an exogenous the economic development of Indonesia variable in the macroeconometric model. since 1969 in terms of macroeconomic The core model was estimated several aggregates. times with data for different sample Indonesia's per capita gross domestic periods. It was first estimated with the product (GDP) was less than 100 dollars data for the sample period 1969 to 1980. in 1970, about 200 dollars in 1974 and Then data for 1981 were added. and it was hit the 300 dollar mark in 1977. Indo re-estimated and revised to give Core nesia became one of the lower middle Model-81. Core Model-81 was used for income economies in the World Bank's planning together with other submodels. classification, attaining a per capita GDP
It was then re-estimated with the data of about 490 dollars in 1980.3) GDP in from 1969 to 1983 to give Core Model-83. creased at an annual rate of about 20 Other submodels were treated similarly, percent in the 11 years from 1970 through and numerical suffixes are added to distin guish the different versions and revisions 2) For detailed explanation of the fiscal sub model and the balance-of-payments submodel, with different sample periods. see Bappenas and Ministry of Finance [1982J This paper aims mainly to describe the and Odano [1983bJ. structure of the core model and to explain 3) The dividing line between low-income and middle-income economies is 410 dollars at its structural equations. Section 2 describes 1982 prices based on the World Bank's classi the historical development of the Indo- fication.
351 Table 1 Comparison with ASEAN Countries
GNP Popula- Average Average Percentage of Labor per capita tion Annual Annual Force in: (1980) (dollars) (millions) Rate of Rate of Growth Inflation (1982) (mid-1982) (%) (%) Primary Secondry Tertiary (1970-1982) (1970-1982) Industry Industry2> Industry
Indonesia 580 152.6 7.7 19.9 58 12 30 Thailand 790 48.5 7.1 9.7 76 9 15 Philippines 820 50.7 6.0 12.8 46 17 37 Malaysia 1,860 14.5 7.7 7.2 50 16 34 Singapore 5,910 2.5 8.5 5.4 2 39 59 Korea. Rep. of 1.910 39.3 8.6 19.3 34 29 37 Hong Kong 5.340 5.2 9.9 8.6 3 57 40
Industrial Market 2.8 9.9 Economies 11,070 722.9° 6 38 56 Japan 10,080 118.4 4.6 6.9 12 39 49 United States 13,160 231. 5 2.7 7.3 2 32 66 Sources: World Bank, World Development Report 1984. Notes: 1) This figure shows total population of industrial market economies. The average population is 38. o. 2) Secondary industry comprises mining, manufacturing, construction and electricity, water, and gas.
1981. due to a large extent to Table 2 Percentage Distribution of Real GDP by Industrial Origin big increases in the price of the crude oil that Indonesia ex I 1970 I 1975 I 1980 I 1983 ported. 1. Primary Industries 45.5 36.8 30.7 29.9 In comparison with ASEAN 2. Mining and Quarrying 10.1 10.9 9.3 7.4 countries other than Brunei, 3. Manufacturing Indonesia had the largest popu 8.4 11.1 15.3 15.6 4. Electricity, Gas, and 1ation and the lowest GDP per Water Supply 0.4 0.5 0.7 0.9 capita in 1982 (Table 1). In 5. Construction 2.8 4.8 5.7 6.3 6. Wholesale and Retail the World Bank's classification, Trade 16.3 17.0 16.6 17.4 Indonesia, Thailand and the 7. Transport and 3.2 4.0 5.5 5.9 Philippines belong to the lower Communication 8. Banking and Other middle-income economies. while Financial Inter- 0.9 1.3 1.9 2.2 mediaries Malaysia and Singapore belong 9. Ownership of to the upper middle-income Dwelling 1.7 2.6 3.0 3.1 economies. Indonesia performed 10. Public Administration and Defence 6.0 7.4 8.7 9.2 as well as other ASEAN countries 11. Services 4.7 3.6 2.8 2.6 in terms of the .rate of growth 12. Gross Domestic Product 100.0 100.0 100.0 100.0 of real GDP between 1970 and
352 S. KURIBAYASHI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia
(%) certain interesting fea 12 Gross Demestic Product 10 tures. First, real GOP 8 follows a conspicuous 6 growth-rate cycle with 4 a ~ _'. two-year up-swing and 2 two-year down-swing. 3g 1--'""19~70;;-::7f:-1 ....,R:b:-~73~74-:-:7:r:5,-:7*6~77:--::!7!:-B -=7:1:-:B*0~8"'-1-:8!:-2....,83~ 9 Second. as can be Construction i \ ...... Manufacturing 24 '. i ...\,-~ guessed from Table 2, 20 "J: \'./\ ! 'v( \ growth rates of GOP and : ~ \ I : \. .:'" primary industry change 10 "'! /\ .: 7 almost in parallel except ',: ".~,j '; Mining \, for 1977. Up to about O~-r------'~------..---:,.---:""...... ,..-+--;..-~-- 1976. in particular. the 24 growth patterns of 20 -10 mainly agriculture and Transportation and Communication mining were reflected in 10 ~. GOP. Third. the .influ ~. Public Administration and Defence ences on GOP of maliu 1970 71 72 73 74 75 76 77 78 79 80 BI 82 83 (year) facturing, construction. Fig.2 Growth Rate of GDP by Industrial Origin and public administra- tion and defence have 1982, but had the highest rate of inflation, increased since 1977, though agriculture 19.9 percent. In the industrial structure and mining were still dominant. Fourth, of employment, Indonesia has high per the real growth rate of all industries drop centage distribution in primary industry ped sharply in 1982, and most industries and low one in secondary industry. except agriculture recorded the lowest rate Table 2 shows changes in the percentage of growth since 1969. Their rates of distribution of real GOP by industrial growth were also very low in 1983. In origin. It is clear that the percentage of building a macroeconometric model, it is primary industry (mainly agriculture) extremely important to assess whether this decreased rapidly while those of manu sharp reduction in the rate of growth of facturing. construction, transportation and real GDP is of a temporary nature or more communication, and public administration permanent. and defence increased. Mining, and To find out why the growth rate of real wholesale and retail trade maintained GOP dropped suddenly at the beginning roughly a constant percentage distribution of the 1980s, we also have to examine throughout the decade. Fig. 2, which GOP by expenditure. Fig. 3, which depicts depicts changes in annual growth rates of growth rates of real GDP by expenditure. real GOP by industrial origin. reveals also reveals a few important facts. First.
353 (%) ing countries turned out to be an 18 r------~ 16 "01'1 bonanza" .Lor IndonesIa.. Income Gross /:\ 14 Domestic was transferred from oil importing 12 Product' ../ \ countries to Indonesia and enabled 10 private consumption to increase at 8 a much higher rate than GDP. This 6 ...... i\ , . 4 .... Private ~. \./ is shown more clearly in Fig. 4. 2 .' Consumption which shows annual movement of o ha::;;--;fI:.~7~3~-7-=!::-:!:;--=!;;~---:!:~-:!-:~ 72 ,...~4 7576 77 78 79 80 81 82 83 the terms of trade of Indonesia. I. 30 . \ The terms of trade substantially im !\ Gross Domestic I. . \ proved in 1974 and in the period 20 :.j..... \ 1979 to 1981:ll That is to say, Indo nesia became able to import three 10 or four times as many goods and services as before the increases in
". the crude oil price with the same ...... ~'•..•.. amount of exports in those years. -10 ...... j V We can also notice that imports 1970 71 72 73 '74 7576 77 78 79 80 81 82 83(Year) increased substantially in the same Fig. S Growth Rate of GDP by Expenditure years. At, the' same time, imports the rate of increase in private final con fluctuated almost in parallel with private sumption fluctuates almost in parallel with <;:onsumption. Second. the rate of increase the growth rate of real GDP, but the former in exports decreased after 1976 and real far exceeds the latter in 1974 and in the exports even decreased after 1979. It was period 1979 to 1981. This phenomenon due to sharp increases in the export may only be observed in oil-exporting price of crude oil that the Indonesian countries. This is attributable to two big economy was able to maintain a high increases in the price of exported crude growth rate in the late 1970s, despite the oil. As Indonesia is an oil-exporting coun deceleration of real exports. try, the so-called "oil shock" for oil import- The sharp drop in the growth rate of real GDP in 1982 was caused mainly by a
(%) steep decrease in exports. which was as 300 cribed to the long-lasting severe world reces
200 Tirms of Trade ..../ 20 sion triggered by the second "oil shock." __L : (.J , 10 100 4) .... Export Price of Crude Oil Though the export price of crude oil de...... -.. creased by five dollars per barrel, the terms of trade in 1983 improved. This is due to in 19697071 7273 74 7S 76 77 78 79 80 '81 8283(year) creases in export prices of primary goods Fig. 0& Terms of Trade and Export because of recovery in the world market for Price of Crude Oil those goods. S. KURIBAYASBI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia
In 1983 non-oil exports of Indonesia in This was caused by the reduction in both creased due to the increase in world trade government consumption and investment. that accompanied the recovery of the world The export price of crude oil was reduced economy. especially the economy of the by five dollars per barrel in March, 1983. United States. The growth rate of real This affected government revenue adversely GDP, however, remained low, though it and aggravated the balance of payments. exceeded the rate in the previous year. In the face of this economic difficulty, the
Table 3 Annual Growth Rate of Real GDP by Expenditure
11969-1981\ 1982 1983 11970-1973\1974-197811979-1983 1. Private Consumption 8.5 3.4 7.5 6.1 7.5 10.9 2. Government Consumption 12.3 8.2 -1.0 15.0 12.3 7.5 3. Total Investment 14.9 13.0 7.8 22.6 14.2 11.1 3.1 Private Investment 9.8 24.0 26.2 18.4 8.9 13.1 3.2 Government Investment 22.6 4.7 -8.5 39.8 23.5 10.1 4. Exports 7.5 -13.9 6.3 16.2 6.3 -3.2 (Non-oil/gas Exports) (4.4) (-) (-) (-) (-) (-) 5. Imports 17.0 8.2 12.3 19.1 15.5 17.0 6. Gross Domestic Product 7.9 2.2 4.2 8.8 7.2 6.1 (5.0*) (6.7*) (6.5*) Note: *The target growth rate in each five-year plan.
Table 4: Annual Growth Rate of Real GDP by Industrial Origin
11969-19811 1982 1983 11970-197311974-197811979-1983 1. Primary Industries 3.8 2.1 4.8 4.6 3.0 4.2 1.1 Agriculture 3.9 3.6 4.8 3.1 3.4 5.5 1.2 Forestry 3.1 -24.1 3.5 20.8 0.4 -10.7 1.3 Fishery 3.6 5.5 6.2 1.1 4.3 5.6 2. Mining and Quarrying 7.5 -':'12.1 1.8 16.7 5.0 -1.7 3. Manufacturing Industries 14.0 1.2 2.2 13.0 13.7 9.7 4. Electricity, Gas, and Water Supply 13.4 17.4 6.9 11. 7 13.5 14.8 5. Construction 16.1 5.2 6.2 23.1 15.2 8.8 6. Wholesale and Retail Trade 8.0 5.7 3.8 10.3 6.5 8.0 7. Transport and Communication 13.4 5.9 5.0 13.3 15.1 7.9 8. Banking and Other Financial Intermediaries 15.4 11. 7 7.0 22.2 14.9 11.0 9. Ownership of Dwelling 14.3 5.2 6.1 15.4 15.1 6.9 10. Public Administration and Defence 11.8 3.6 5.5 7.9 13.9 9.1 11. Services 2.5 2.3 2.5 2.6 2.4 2.4 12. Gross Domestic Product 7.9 2.2 4.2 8.8 7.2 6.1
365 Government restrained' its expenditure sUPply attained accelerating .. growth rates severely and devalued foreign exchange. through ·the three periods, though the In this connection, real. government con acceleration was not large. sumption .decreased. by 1. 0 percent and On the .expenditure 'side, -the rate of real government investment by 8. 5 per increase in exports dropped sharply, while cent (Table 3). private consumption increased at an It is crucial to recogniie that Indo accelerating rate. This was attributed to nesia's external economic environment the "oil bonanza." became much less favorable in the early Indonesia has given priority to agri 1980s than in the 1970s. The five-dollar cultural development to attain self-suffi reduction in the oil export price was ciency in food and to developing and literally a "reve'rse oil shock" to Indonesia. fostering import substitution industries This is one of ,key ele~ents of our model mainly in the field of consumption goods. building. One ·problem is that the development of Tables 3 and 4 show respectively the import substitution industries did. not trends of real GDP by· expenditure and necessarily lead to export promotion and by industrial origin. Both tables are based Indonesia lagged far behind in developing on tlie assumption that growth trends export industries. Exports. therefore, 'con changed in the. early 1980s. The last three sist mainly of crude oil and primary pro columns in both tables show annual average ducts. In the 1970s crude oil export growth rates for the past three five-year brought in enough foreign currency for development plans. Observed growth Indonesia to import both consumer goods rates of real GDP exceeded the targeted and capital goods needed to pursue its growth rates for Repelita I and Repelita development policies. After 1982, however, II, but fell short of those for Repelita III Indonesia encountered difficulties in earn because of the low growth rates in 1982 ing foreign currency and had to employ and 1983. Compared with the previous austerity policies. Another important fea two planning periods, the average annual ture of the Indonesian economy is that rates of increase in 1979-1983 decreased government expenditure, both consumption for m'ost items in both tables. This may and investment, increased at the highest mean ., that as'the scale of the Indonesian rate in the 1970s and led the economy. economy increased, the economic·domestic This is a common feature of almost all fronti~rs were reduced and economic deve developing countries. Since Indonesia lopment decelerated. The rates of increase depends on crude oil export--about 60 in real value added of forestry and mining percent of total export is crude oil and dropped sharply, as did the rate of increase about '70 percent of government _domestic in construction, although this still remained revenues is related to oil production -and relatively high..On the other hand, fishery, export-'-decreases in the export price of agrictl.1t.u!"~, _~md_ electri<::ity, gas. and water crude oil and the quantity of crude oil
356 s. KURIBAYABSI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia export compelled the Government to impossible, for developing countries, espe reassess its economic policies and develop cially low-income economies and lower ment strategies in the early 1980s. middle-income economies, to finance their economic development with only domestic 2. Economic Problems in the 1980s saving, for their saving rates are very low. We have seen that Indonesia was con Consequently, they have to earn foreign fronted with several economic problems currency either by exports or by overseas in the early 1980s. These have to be taken financing such as foreign aid, borrowing into account in constructing a macroecono or direct investment, or by both. metric model. Some of them will be touched In the 1970s Indonesia earned enough upon here. foreign currency for her rather high 2.1 Reduction ofthe Excessive Dependence economic development by exporting crude on Petroleum and LNG oil and gas. During the same period the Indonesia depends on crude oil export Indonesian Govenment was able gradually in two aspects: to earn the foreign currency to reduce the ratio of foreign aid to total necessary for economic development, and government revenue, while increasing the to finance government expenditure, espe share of development expenditure in total cially government development expenditure. expenditure. Indonesia was, however, Prospects for the quantity and price of confronted with difficulties in increasing crude oil export, therefore, are crucial crude oil export in the face of "the reverse factors in the future economic develop oil shock" of the early 1980s. High priority, ment of Indonesia. This means that the therefore, needs to be given to the pro Indonesian economy is substantially influ motion of non-oil exports in place of enced by the world economic conditions. crude oil and gas in Repelita IV. On the Indonesia, therefore. has to reduce its other hand. one way to save foreign excessive dependence on crude oil in order currency is to foster and develop import to attain a stable economic growth rate, substitution industries so as to reduce by promoting non-oil exports on one hand imports and improve the balance of pay and on the other by changing government ments. Indonesia has adopted this devel revenue systems so as to increase tax opment strategy for consumption goods revenue from sources other than oil com industries and obtained good results. One panies and by reducing subsidies on re problem is that the development of import fined petroleum products. substitution industries did not necessarily 2.2 How to Finance the Economic Devel lead to the development and promotion of opment export industries. Another policy measure Needless to say, developing countries for increasing exports and reducing imports need foreign as well as domestic capital is devaluation of the exchange rate, which in order to achieve adequate economic has been done several times in the past. development. It is usually difficult, if not In the past, however, devaluation of the
367 rupiah contributed to the reduction of Table 1 showed that the rate of infla imports but not to the promotion of ex tion in Indonesia in the period 1970 to ports. 1982 was very high compared with other Capital goods industries have not yet countries. Indonesia has been trying to been developed in Indonesia, so that in increase domestic savings by changing creases in public and private investment its financial systems. For this purpose, serve to increase imports and worsen the among others, it is crucial to control infla balance of payments, as will be shown tion in Repelita IV. later by multiplier analysis. It is a diffi cult and challenging task for policy III Medium-term Macroeconometric makers to decide what strategies should Model for Economic Planning be adopted for developing capital goods industries. Different strategies from those 1. Basic Framework of Macroeconometric adopted for consumption goods industries Models might be considered. for their impacts on The major role played by macroecono others sectors will be more profound and metric models in economic development far-reaching. planning is that of tracing and estimating 2.3 Creation of Employment Opportunity the relationship between economic policy It is estimated that 1. 5 to 2. °million objectives and policy measures in the mainly young people will enter the labor light of past economic development, and market every year in the coming years. of selecting optimal policy measures for An important task is to create jobs for this achieving the policy targets set in order young labor force. According to the input to solve the problems in the economy. output table for 1980, the percentage Our macroeconometric models were, distribution of employment among indus therefore, designed to assess quantita tries is as follows: primary industry, 54. 1; tively the impact of government economic manufacturing, 9. 6; construction, 3. 0; policies on a variety of macroeconomic wholesale and retail trade, 12. 4; transport aggregates, especially target variables. As and communication, 2. 8; services and mentioned, world economic conditions in others, 16.3. It is also well known that ~uence the Indonesian economy to a large most of the labor force is employed in extent. so that variables representing world small-scale or cottage industries, especially economic conditions also have to be incor those using domestic resources. In the porated into the macroeconometric models face of an decelerating economic growth to, estimate their impact on the Indonesian rate, an economic and social system must economy. Fig. 5 shows major policy varia be devised to make effective use of domes bles, data variables and target variables. tic resources and to absorb new entrants Main policy variables belong to the to the labor force. government sector and are closely con '2.4 Control of Inflation nected with the fiscal submodel. As far
SI8 S. KURIBAYASBI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia
Major Policy Variables Target Variables Government Consumption (CG) Growth Rate of Real GDP (GDPR) Government Investment (IG) Per capita National Income Various Tax Parameters M (NNPIN or GDPIN) Price of Refined Oil for Domestic Inflation Rate (PGDP or PCP) Required Foreign Financing Consumption (PDROL) o Money Supply (SMB) Required Government Foreign Aid· Credit Supply to Private Sector (CRPMS) Current Deficit in the Balance Exchange Rate (RFEX) D of Payments Labor Balance (LABF and EMP)
Major Data Variables Oil Balance World Imports (MWR) E Quantity of Oil Production (QOIL) --\ Export Price of Crude Oil (PXOIL) --; Other Endogenous Variables L Quantity of Export of LNG (QXGAS) Export Price of LNG (PXGAS) Population (N)
Fig.5 Policy and Target Variables in the BAPPENAS Macroeconometric Models as the government revenue is concerned, petroleum products in the 1970s to keep tax parameters are considered as policy their domestic prices much lower than parameters. Although various tax equa- international market prices of their tions were estimated for the sample pe equivalents. Facing economic difficulties riods, the parameters of tax equations may at the beginning of the 1980s, the be changed so as to estimate the impact Government introduced new policies to of tax reform on government revenue and reduce the subsidies and to raise the other economic aggregates during the prices to international levels during Repe planning period. Important policy varia lita IV. To assess the government policy, bles in our models are government con the price of refined oil for domestic con sumption, which roughly corresponds to sumption was introduced into the core the central government's routine expendi model as a policy variable. We also ture, and government investment, which sought to introduce a government-regu corresponds to the cental government's lated price index, made of prices of com development expenditure less defence ex- modities such as rice and sugar, in order penditure and subsidies on fertilizers. to assess the impact of government price The Indonesian Government subsidized policy on the rate of inflation and other
369 variables. But lack of data prevented us there are two approaches often used in from making an adequate price index, constructing macroeconomic models for which should be constructed in the future. developing countries. One emphasizes the Money supply, which is the nominal supply side, based on the assumption that supply of broad money in our model, and a shortage of the production capacity of the loans to the private sector, which is goods and services puts a ceiling on the the amount of credit supplied to the economic growth of developing countries. private sector by the monetary system, We call this "the supply-side approach." are introduced into our model as govern The other emphasizes the demand side, ment monetary policy variables. As men based on the assumption that a lack of tioned later, we had difficulty in incor demand restricts the economic growth of porating these variables into the core developing countries as well as developed model. Exchange rate is one of the most countries. We call this "the demand-side important policy variables. approach." Ideally, both sides should be The other exogenous variables which taken into account simultaneously, for in play an influential role in the macroecono reality the production and demand sides metric model are called here "data varia interact. Emphasis, however, has to be bles." World imports, quantity of oil pro laid on one side or the other in a small duction, and export price of crude oil econometric model with less than one influence crucially the future course of hundred equations. A lack of data also the Indonesian economy as well as govern compels us to select one approach. Both ment development policies. Population is approaches were tried at the beginning treated as a data variable, though it will of our model building. be a target variable in the long run with First, the supply-side approach was the promotion of family planning. pursued in the core model. A macro The target variables listed in Fig. 5 are production function of the Cobb-Douglas among those required for the economic type was introduced, and private con indicators in the five-year development sumption was to be estimated as a resid plan. Such endogenous variables as cur ual. But various difficulties were encoun rent balance, rate of inflation and the ratio tered in applying the supply-side approach. of foreign aid to government revenue are One was a lack of data on capital stock considered as constraints to the govern and employment. With this approach, ment policies. The Government pursues sectoral production functions should be optimal economic development policies estimated. But this was impossible be subject to the constraints. cause no data on sectoral capital stock Once endogenous and exogenous varia and employmetnt were available. Even on bles are determined, structural equations the macro basis, no data on gross capital have to be specified and estimated with stock were available. Another difficulty data for a large enough sample. Broadly, was that the core model based .on this
360 S. KURIBAYASHI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia approach could not predict the economic 2. Structual Equations of the Core Model slowdowns which occurred in 1982 and The core model consists of 59 equa 1983. This model also had difficulty pre tions: 21 structural equations and 38 iden dicting up-swings and down-swings within tities. The fiscal submodel includes 12 the sample period. For these reasons, the tax-and-revenue-sharing equations and one demand-side approach was adopted at the identity. But only the system of equations final stage. for the core model is shown in the following Second, a private consumption function "System of Equations." The structural was introduced, and the production func equations of the core model will be ex tion was modified slightly and used in a plained here. Those of the fiscal submodel different functional form for estimating are explained in Bappenas and Ministry of employment. That is to say, the employ Finance [1982]. Equation numbers in the ment function was specified based on the following explanations refer to those in the assumption that employment was deter system of equations. mined on the production function, given 2. 1 Private Consumption Function (Equa the volume of production and capital stock. tion 2) The core model based on the demand The private consumption function in side approach showed much better perfor cludes real private national disposable in mances in the interpolation and extrapo come, real money supply and real private lation tests and could trace the declines consumption with a one-year time-lag as in growth rate in 1982 and 1983. explanatory variables. Since Indonesian Third, the supply-side approach was also national accounts do not include the in applied to the export functions. Exports come and outlay accounts of the house were divided into three categories: crude hold sector, data on household or personal oil, gas, and non-oil and non-gas (non-oil disposable income are not available. The exports for short) . In specifying each first explanatory variable, (NNP-TY) jPCP, export function, export was essentially is a proxy variable for real household estimated as the difference between domes disposable income. The second explanatory tic production and domestic consumption. variable, 5MB(-l)jPCP(-l), which is Good results, however, were not obtained real money supply at the end of the pre for non-oil exports, so that a non-oil vious year, is introduced to represent the exports function was introduced. As far real balance effect, through which money as crude oil export was concerned, the supply influences other endogenous varia supply-side approach was adopted in the bles. Although significant estimates of its final version. coefficient were not always obtained, this After scrutinizing the results of both explanatory variable was included because approaches, the demand-side approach the inclusion of money supply in the core was selected for all functions except those model would have become meaningless related to crude oil. without money supply in the consumption
361 the private consumption Rate of Rate of Increase Increase in Year function in Core Model-81 in CPR (NNP-TY) /PCP -4 0 4 10 20 30 (%) in order to treat the private (%) (%) consumption observation for 7.6 3.0 1970 1981 as an irregular move 4.7 6.5 1971 5.8 12.8 1972 ment. II. 1 13.0 1973 This irregularity may be 14.5 18.3 1974 caused by the compilation 3.6 2.7 1975 method of private conSump 8.0 10.6 1976 4.0 7.5 1977 tion in the national accounts 7.5 5.2 1978 of Indonesia. Private coil- ...,:....., 14.3 24.7 1979 ..... sumption, nominal and real, 12.7 14.7 1980 is compiled as a residual 16.7 8.8 1981 between the sum of gross '3.4 -3.2 1982 7.5 7.9 1983 domestic product and im- Fig.6 Rates of Increase in Real Private Consumption ports and the sum of final and Real Private National Disposable Income expenditures other than private consumption and function. Real private consumption with increases in inventory. one-year time-lag, CPR ( -1), is included According to our estimates, the long based on a distributed lag of the Koyck term prospensity to consume with respect type. to private national disposable income was When the private consumption function O. 613 in Core Model-81 and 0.655 in Core was reestimated with the data extended Model-81 and 0.655 in Core Model-83. up to 1981, the opposite sign was obtained 2.2 Real Gross Domestic Private Fixed for the coefficient of real money supply. This Capital Formation (Equation 4) seemed to be caused by an extremely high The stock adjustment principle was rate of increase in real private consump employed mainly to provide explanatory tion relative to real private national dis variables for real gross domestic private posable income in 1981. As Fig. 6 shows, fixed capital formation (IPR, real private the rate of increase in real private con investment for short). Real private invest sumption is extremely high, 16. 7 percent, ment was regressed on real gross domestic despite the relatively low rate of increase, product with one-year lag (GDPR (-1», 8. 8 percent, in real private national dis net capital stock at the end of the pre posable income. This type of movement vious year (KPR ( -1) ), and increase in was not observed over the sample period real credit supply to private sector by 1969 to 1980, although the same phe monetary system (CRPMS/PI). Other in nomenon took place in 1982. A dummy vestment theories such as the neoclassical variable was, therefore, introduced into type and Tobin's q could not be tried be-
382 S. KURIBAYASHI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia cause of lack of data. consumption QDROL is converted into the Bank loans to the private sector in quantity of crude oil by equation 9, which fluence other endogenous variables through follows the technical relation for conversion. this equation. although a significant esti (2) Function of the Quantity of Refined mate of its coefficient was not obtained. Oil for Domestic Consumption (Equa 2. 3 Exports Functions (Equations 6 tion 10) through 13. 35 through 40) The quantity of refined oil for domestic Exports of goods and services are disag consumption (QDROL) is a log-linear func gregated into three components: oil. gas, tion of real total consumption (CR), real and non-oil and non-gas exports. Exports total investment (IR) and the ratio of price of goods and services in the national of refined oil for domestic consumption to accounts of Indonesia are not divided. consumer price index (PDROL( -1)/PCPI and data on these components had to be ( -1)). The first two explanatory variables sought from other sources. The sum of represent income effect and the last one these components. therefore, is not neces represents price effect. Though real GDP sarily equal to exports of goods and ser was tried as representative of income vices in the national accounts, but includes effect in place of CR and IR, better results statistical discrepancies. Consequently, the were not obtained. The difference in statistical equations 13 and 39 were esti elasticities with respect to consumption mated in order to complete the model. and investment are considered to be rea 2. 3. 1 Crude Oil Export and Quantity sonable. A reasonable and more-or-Iess of Refined Oil for Domestic Use stable estimate of price elasticity was also (Equations 6 through 10) obtained as shown in the system of equa As mentioned, the quantity of crude oil tions. export (QXOIL) is obtained as the differ It will be desirable in the future version ence between production and domestic to break down refined oil products into use, as shown by equation 8. main petroleum products such as kerosene. (1) Crude Oil Export Identity (Equation 2.3.2 Real Gas Export (Equation 11) 8) Gas export (XGASR) is essentially de QXOIL= QOIL# + QMOIL# - QDOIL termined exogenously. Almost all of gas + QXOSD#. export is LNG. and the quantity is where QOIL denotes the quantity of determined by the long-term production crude oil production. QMOIL the and export contracts. The export price of quantity of crude oil import. QDOIL gas (PXGAS) follows that of crude oil, the quantity of crude oil for domestic which is an exogenous variable. consumption and QXOSD statistical 2.3.3 Real Non-oil and Non-gas Exports discrepancy, and the symbol # indi Function (Equation 12) cates that the variables are exogenous. The real non-oil and non-gas exports .The quantity of refined oil for domestic (XNOSR) function was specified in accord-
363 ance with the typical theory of demand imports of consumption goods, as· will be function. The explanatory variables consist shown in the multiplier analysis. mainly of real world imports (MWR) as the 2.4.2 Function of Imports of Investment relevant income variable and the ratio of Goods (Equation 15) price of non-oil exports (PXNOS) to world The long-term elasticity of imports of export price (PWX) as the relative price investment goods with respect to real in variable. One important channel through vestment in Core Model-81 is O. 945, much which world economic conditions influence higher than that of imports of consumption other endogenous variables is this equation. goods. The difference between the elastic A dummy variable was introduced to ex ities is reflected in the difference between plain unusually sharp declines in 1982 and the multipliers of government consumption 1983, without which the equation did not and investment, as will be shown later. In fit well. One reason for the sharp decline Core Model-83, the estimated long-term was that Indonesia banned the export of income elasticity is 1. 014, slightly larger lumber and changed its policy for shrimp than 1. 0, although it should be less than export. It should be noted that the esti or equal to 1. O. These estimates of coeffi mated price elasticity is very low. cients indicate that almost all investments 2.4 Import Functions (Equations 14 have had to be imported because capital through 18, Identities 44 through 48) goods industries have not yet been devel Imports of goods and services are disag oped in Indonesia. gregated into four components: consump 2.4.3 Function of Imports of Raw Mate tion goods, investment goods. raw materi rials and IntermediateGoods (Equa als and intermediate goods, and services tion 16) and statistical discrepancy. Like the func Real imports of raw materials and inter tion of non-oil exports, all imports were mediate goods were specified as a function regressed on a variable representing in of real GDP (GDPR) and the ratio of im come effect and a variable representing port deflator for raw materials and inter price effect. mediate goods to GDP deflator (PMRM/ 2. 4. 1 Function of Imports of Consump PGDP). Fairly stable estimates of the tion Goods (Equation 14) coefficients were obtained. The fit also Imports of consumption goods (MCR) improved as more recent data were added are a function of real total consumption to the sample. (CR) and the ratio of import deflator for 2. 4. 4 Function of Imports of Services consumption goods to consumption deflator and Statistical Discrepancy (Equa (PMC/PC). A fairly high estimate of price tion 17) elasticity of consumption-goods-imports Satisfactory results. were not always ob demand was obtained. This means that tained in terms of fit, although a stable coeffi the government policy toward foreign ex cient was estimated for each version. This change can exert a strong influence on is due to a defect in the way the data
3M s. KURIBAYASHI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia were compiled. The data on imports of ser to the desired level within one year due to vices and statistical discrepancy (MSDR) various costs involved. So they are assumed were compiled as the difference between to adjust their employment partially. This real imports of goods and services (MR) is formulated as follows, in the national accounts and the sum of (EMP/EMP(-1» =A (EMp·/EMP(-1)), MeR, MIR, and MRMR, which were where O<.«l. obtained from other data sources. MSDR, Substituting EMP* into this equation and therefore, includes not only imports of taking logarithms of both sides gives the services but also statistical discrepancy. following labor demand function: The problem is that MSDR accounts for lnEMP= ao.< +a1AlnGDPR - az.
365 tor in accordance with some theory, then in 1981 and 1982. In the 83 version. to specify other deflators as a function of unfortunately, a significant estimate was GDP deflator and other variables. The not obtained for the coefficient of (GDPR/ other is first to estimate individual deflator EMP). functions, then to obtain GDP deflator as the ratio of nominal GDP to real GDP. In PCP ICore MOdel-Sllcore Model-83 The latter approach is adopted in most -19.4192 -15.6669 Canst. other macroeconometric models. (- 3.22) (- 8.85) According to the former, GDP deflator 1. 9533 1. 7698 In CPR plays a central and crucial role in price (3.61) (6.92) determination. First, the monetary approach In( GDPR ) - 1.0475 was applied to the specification of the EMP (- 1.04) GDP deflator function in the core model. In PDROL 0.3339 0.1323 Good results were not obtained, however, (-1) (1.60) (0.89) especially for the extrapolation. Conse - 0.373625 - 0.2294 DUMPC quently, the latter approach was adopted in (- 4. 01) (- 4.01) the final version of the core model. In this 2 case, private consumption deflator and R: 0.986 0.990 investment deflator are essential. In each function the shift parameters of demand 2. 8. 2 Consumer Price Index Function and supply functions are basically used as (Equation 57) explanatory variables, for prices are as Consumer price index (PCPI) was simply sumed to be determined at the intersection regressed on PCP, as both are of the of demand and supply schedules. same kind. PCP covers all goods and GDP deflator is determined as an im services, whereas the coverage of consumer plicit deflator by equation 5l. price index is limited. The main differ 2. 8. 1 Private Consumption Deflator ence theoretically is that consumer price Function (Equation 25) index is of the Laspyres type and PCP is Private consumption deflator (PCP) is of the Paasche type. a function of real private consumption 2. 8. 3 Government Consumption Deflator (CPR), labor productivity (GDPR/EMP), Function (Equation 26) and price of refined oil for domestic Government consumption consists mainly consumption (PDROL) in Core Model of the compensation of government employ 81. As mentioned, we sought to intro ees and the goods and services which duce a government-regulated price index the Government purchases. Government into this function. PDROL was introduced consumption deflator (PCG) , therefore, in place of the price index. A dummy should be a function of the wage rate of variable was used to eliminate the effect government employees and the price of of the irregularity of private consumption the goods and services. But as the wage
368 S. KURIBAYASHI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia rate was not available, PCG was regressed in the fiscal submodel. on a kind of domestic demand deflator (CP+I) / (CPR+IR) = PCP*CPR/ (CPR + 3. System of Equations IR) +PI*IR/(CPR+IR). Notes: 1. The symbol # denotes exogenous 2. 8. 4 Investment Deflator Function variables. (Equation 27) 2. The first figure in parentheses Investment deflator (PI) is a function of after an equation name shows labor productivity, money supply (SMB) the equation number in Indo and import deflator of investment goods nesia's Economic Development (PM!). We tried to introduce real total and Bappenas Macroeconometric investment (IR) in place of 5MB, but a Model, 1969-1980by K. Kobayashi significant coefficient of productivity could [1982], and the second figure not be obtained. is its number in Model Makroe Because most investment goods are im konometri Inti Indonesia by ported, PMI is assumed to be a dominant Bappenas [1982]. factor for PI, but the assumption was not 3. RR denotes the coefficient of fully supported by our results. This may determination adjusted for de be ascribed to construction investment. grees of freedom, s the standard 2.9 Production Function and Capacity deviation of disturbances and Output (Equations 1 and 58) DW Durbin-Watson. ratio. The production function of the Cobb 4. The figures in parenthess below Douglas type was estimated. The capacity coefficients are t-values of the output equation was derived from the pro coefficients. duction function by shifting the production 5. The sample period is from 1969 function upward by the amount of maxi through 1983. mum residual within the sample period. 1. Real Gross Domestic Product (1, 1) (For details, see appendix B in Kuribayashi In (GDPR/EMP) = -0. 885399 [1982].) (-14.67) Capacity output was used in our model + O. 6830 In (KR ( -1)/EMP) to compare real GDP with capacity output (13.94) (GDPRC). GDPR/GDPRC ratio more or RR=O.937 s=0.0461 DW=0.47 less expresses a kind of capacity utilization. For Model I 2. 10 Tax Function (Equations 53 and 56) or Two tax functions were estimated in GDPR=CR+IR+XR-.MR the core model: a net indirect tax equa For Model II tion and a direct income tax equation. 2. Real Private Consumption Expenditure When the core model and the fiscal sub (l0,14) model are integrated, these equations are CPR=GDPR- (CGR+IR+XR-MR) replaced by the corresponding equations For Model I
367 or +0.6686 InIR CPR = 392. 24 + O. 3082 (NNP-TY)/PCP (6.85) (0.85) (3. 60) -0.2672 In(PDROL( -1)/PCPI .+0. 3475 (SMBI+SMB#(-1»/2/ (6.22) (0.52) ( -1» PCP+ O. 5296CPR( -1) RR=0.997 s=0.0278 DW =2.43 (3.29) 11. Real Export of Gas (19,30) PR=0.996 s=170.4 DW=2.04 XGASR= (XGAS$I*415. O)/PXGASI*I/ For Model II 1000 3. Real Consumption Expenditure (9, 15) 12. Real Non-oil and Non-gas Exports (20, CR=CPR+CGR# 32) 4. Real Gross Domestic Private Fixed InXNOSR = O. 258415 + O. 9814 InMWRI Capital Formation (12, 23) (0. 51) (12. 3) InIPR= -6.0734+2. 34111nGDPR(-1) -0.1129 In (PXNOSI (-1) /PWXI (2. 26) (1. 41) (-2.43) - O. 9869 InKPR ( -1) (-1» (-0.68) - O. 2451DUM8182 + O. 0609 In (.::1CRPMS/PI) (-10.44) (0. 71) RR=0.953 s=0.0303 DW=2.62 RR=0.921 s=0.117 DW = 1. 26 13. Real Total Export (13. 34) 5. Real Gross Domestic Fixed Capital XR=166. 24+0. 9422 (XOILR+XGASR Formation (11, 24) (1. 44) (11. 29) IR= IPR+ IGRI +XNOSR) 6. Real Export of Crude Oil (14, 31) RR=0.900 s=113.26 DW=1.17 XOILR = (XOIL$*415. 0) / (PXOILI/ 14. Real Imports of Consumption Goods 4.0094)*1/1000 (23,41) 7. Value of Oil Export in US $(15,30) InMCR = 2. 75952 -1. 7046 In (PMC/PC) XOIL$=PXOILI.QXOIL (1. 42) (4. 88) 8. Quantity of Crude Oil Export (16, 29) +0.2892 InCR QXOIL = QOILI + QMOILI - QDOIL (1. 33) + QXOSDI RR=0.873 s=0.189 DW=1. 43 9. Quantity of Crude Oil for Domestic 15. Real Imports of Investment Goods (24, Use (17,28) 42) QDOIL=1/0. 7* (1/158. 99*QDROL InMIR = - O. 7968 - O. 9797 In (PMI/PI) -QMROLI) ( -1. 09) (3. 27) 10. Quantity of Refined Oil for Domestic + O. 7534 InIR + O. 2568 InMIR( -1) Use (18,27) (3. 50) (1. 45) InQDROL=1.10478+0.3542InCR RR=0.919 s=0.141 DW =0.94 (2. 02) (2. 54) 16. Real Imports of Raw Materials and
368 s. KURIBAYASHI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia
Intermediate Goods (25, 43) +0.6810 InEMP(-1) MRMR = -120. 56 -159. 2 (PMRM/ (4.00) (-0.75) (-1. 68) RR=0.996 s=0.00786 DW=1.52 PGDP) + o. 0984 GDPR For Model II (10.43) 22. Labor Force (5, 11) RR=0.984 s=48.32 DW=2.01 LABF= -26331. 6+0. 5635N# 17. Real Imports of Services and Real ( -18. 11) (52. 38) Statistical Discrepancy (26, 44) RR=0.995 s=550.85 DW=0.40 InMSDR= -17. 0284+2. 6458 InGDPR 23. Unemployment (6,12) (- 6. 99) (9. 79) UNEM= LABF - EMP RR=0.871 s=0.3289 DW=1. 34 24. Real Gross National Product (2, 2) 18. Real Total Import (22, 45) GNPR=GDPR+NFIAR# MR = MCR +MIR+MRMR + MSDR 25. Private Consumption Deflator (32, 16) 19. Real Depreciation (8, 3) InPCP = -15. 6669 + 1. 7698 InCPR (1) DEPR= -745. 65+1216. 97(GDPR/ (-8.85) (6.92) (-4.44) (5.33) +0.1323 InPDROL( -1) +0.2294 KR(-1) +0. 0388 KR( -1) (0.89) (-4.01) (18.17) DUM 7080 RR=0.987 s=19.49 DM=0.57 RR=0.990 s=0.0641 DW =1. 28 (2) DEPPR= -1225.18+ 564. 222 (GDPR/ 26. Government Consumption Deflator (29, (2. 70) (2. 55) 17) KPR( -1» +0.1685 KPR( -1) PCG= -0. 0642 + 1. 0396 «CP+I) j (CPR (13.14) ( -1. 04) (70. 03) RR=0.982 s=32.34 DW =2.03 +IR» 20. Real Capital Stock (7, 13) RR=0.997 s=0.0684 DW=2.26 (1) KR=KR( -1) +IR-DEPR 27. Investment Deflator (30,25) (2) KPR=KPR( -1) +IPR-DEPPR InPI= -4. 4014-0. 4412 In(GDPRjEMP) 21. Employment (4, 10) (-2.27) (-0.69) InEMP=3. 22128+0.1702 InGDPR +0.5201 InSMB#+O.1351 InPMI (2. 33) (2. 19) (4.41) (1. 45) + O. 5587 InEMP (-1) RR=0.994 s=0.0495 DW=l. 57 (2.89) 28. Nominal Private Consumption Expen RR=0.994 s=0.00938 DW=0.84 diture (44 or 31, 20) For Model I CP=PCP.CPR or 29. Nominal Government Consumption Ex InEMP=2. 21448+0.2273 InGDPR penditure (45,21) (1. 80) (3. 28) CG=PCG.CGR - O. 0862 InKR ( -1) 30. Nominal Total Consumption Expendi (2.39) ture (43, 22)
389 C=CP+CG MC=PMC*MCR 31. Total Consumption Deflator (, 18) 45. Nominal Imports of Investment Goods PC=C/CR (53.51) 32. Nominal Gross Domestic Private Fixed MI=PMI*MIR Capital Formation 46. Nominal Imports of Raw Materials and IP=PI*IPR Intermediate Goods (54,52) 33. Nominal Government Investment MRM= PMRM*MRMR IG=PI*lGR 47. Nominal Imports of Services and Sta 34. Nominal Gross Domestic Fixed Capital tistical Discrepancy (55, 53) Formation (46, 26) MSD = PMSDI*MSDR I=PI*IR (=IP+IG) 48. Nominal Total Import (51, 54) 35. Nominal Oil Export (48,36) M=MC+MI+MRM+MSD XOIL= XOIL$*RFEXI*I/1000 49. Import Deflator (35, 49) 36. Nominal Gas Export (49.39) PM=M/MR· XGAS = XGASI*RFEXI*I/1000 50. Nominal Gross Domestic Product (39, 37. Nominal Non..oil and Non-gas Exports 6) (50.38) GDP=C+I+X-M XNOS = XNOS$*RFEXI*1/1000 51. Deflator for Gross Domestic Product 38. Value of Non-oil and Non-gas Exports (28,5). in US $ (21, 37) PGDP=GDP/GDPR XNOS$=XNOSR*PXNOSI*1000/415. 0 52. Nominal Gross National Product (40, 39. Nominal Total Export (47,40) 7) X= -74. 40+1.0489 (XOIL+XGAS GNP=GDP+PM*NFIAR ( -1. 01) (115. 22) 53. Nominal Net Indirect Tax +XNOS) TI=141. 46+0. 03196 GDP RR=0.999 5==196.9 DW=1.42 (3. 89) (28. 79) 40. Export Deflator (34, 35) RR=0.983 s=94.85 DW=l. 20 PX=X/XR 54. Nominal Depreciation (42, 8) 41. Import Deflator for Consumption Goods DEP= PGDP*DEPR (36,46) 55. Nominal Net National Product (41, 9) PMC= PMC$I*RFEX#/415. 0 NNP=GNP-DEP-TI 42. Import Deflator for Investment Goods 56. Direct Income Tax (37.47) TY= -47. 66+0. 02871 NNP PMI= PMI$#*RFEX#/415. O· ( -1. 88) (32. 03) 43. Import Deflator for Raw Materials and RR=0.987 s=66.11 DW=1. 27 Intermediate Goods (38, 48) 57. Consumer Price Index (33,19) PMRM= PMRM$#*RFEXI/415. 0 InPCPI = 3. 85035 + 1. 02984 InPCP 44. Nominal Imports of Consumption Goods (208.66) (45.22) (52.50) RR=0.993 s=0.0573 DW=0.93
870 S. KURIBAYASHI: A Medium-tenn Macroeconometric Model for Economic Planning in Indonesia
58. Capacity Output duct In(GDPRCjEMP) = -0. 86840 GDPR = Real Gross DomestLc Product +0.6830 In(KR( -l)jEMP) GDPRC = Real Capacity Output 59. Real Net Factor Income from Abroad GNP = Nominal Gross National Pro NFIAR= -68.11-0.1884(XOIL+XGAS) duct (-2.32) (-14.29) GNPR = Real Gross National Product jPM I = Nominal Gross Domestic Fixed RR=0.936 s=65.0 DW=1. 46 Capital Formation IG = Nominal Gross Government 4. Notation of the Variables Fixed Capital Formation Notes: 1. The symbol # denotes exogenous IGR# = Real Gross Government Fixed variables. Capital Formation 2. Real = 1973 constant price. IP = Nominal Gross Domestic Pri Nominal = current price. vate Fixed Capital Formation C = Nominal Consumption Expen IPR = Real Gross Domestic Private diture Fixed Capital Formation CG = Nominal Government Consump IR = Real Gross Domestic Fixed tion Expenditure Capital Formation CGR# = Real Government Consumption KPR = Real Private Capital Stock Expenditure KR = Real Total Capital Stock CP = Nominal Private Consumption LABF = Total Labor Force Expenditure M =Nominal Total Import CPR = Real Private Consumption Ex MC = Nominal Imports of Consump penditure tion Goods CR = Real Consumption Expenditure MCR = Real Imports of Consumption CRPMS# = Amount of Credit Supply to Goods Private Sector by Monetary MI =Nominal Imports of Investment System Goods DEP = Nominal Depreciation MIR = Real Imports of Investment DEPPR = Real Private Depreciation Goods DEPR = Real Depreciation MR = Real Total Import DUM7080=Dummy Variable for Private MRM =Nominal Imports of Raw Mate Consumption Deflator (1 for rials and Intermediate Goods 1970-1980, 0 for 1980-1983) MRMR = Real Imports of Raw Materials DUM8182=Dummy Variable for Non-oil and Intermediate Goods and Non-gas Exports (1 for MSD = Nominal Statistical Discrepancy 1981-1982, 0 otherwise) for Import Sector EMP =Total Employment MSDR = Real Statistical Discrepancy for GDP = Nominal Gross Domestic Pro- Import Sector
371 MWR# =Real World Imports PXOIL# =Export Price of Crude Oil in N# = Population US $ per barrel NFlA = Nominal Net Factor Income PWX# = World Export Price from Abroad QDOIL = Quantity of Crude Oil for Do NFIAR = Real Net Factor Income from mestic Consumption in million Abroad barrels NNP =Nominal Net National Product QDROL =Quantity of Refined Oil for NNPR =Real Net National Product Domestic Consumption in mil PC = Consumption Deflator lion liters PCG = Government Consumption QMOIL# =Quantity of Crude Oil Import Deflator in million barrels PCP = Private Consumption Deflator QMROL# =Quantity of Refined Oil Import PCPI = Consumer Price Index in million barrels PDROL# =Price of Refined Oil for Dome- QOIL# = Quantity of Oil Production stic Consumption QXGAS . =Quantity of Export of LNG PGDP = GOP Deflator QXOIL = Quantity of Crude Oil Export PI = Fixed Capital Formation De in million barrels flator QXOSD# = Statistical Discrepancy for the PM = Import Deflator Quantity of Oil Export PMC = Import Deflator for Consump RFEX# = Rate of Foreign Exchange tion Goods 5MB# = Nominal Supply of Broad Money PMC$# = Dollar Price Index for Con TI =Nominal Net Indirect Tax sumption Goods Imports TIME# =Time Trend PMI = Import Deflator for Investment TIR =Real Net Indirect Tax Goods TY = Direct Income Tax PMI$# = Dollar Price Index for Invest- UNEM = Unemployment ment Goods Imports X = Nominal Total Export PMRM = Import Deflator for Raw Mate XGAS = Nominal Value of Gas Export rials and Intermediate Goods in billion Rp PMRM$# = Dollar Price Index for Raw XGAS$# =Nominal Value of Gas Export Materials and Intermediate in million US $ Goods Imports XGASR =Real Gas Export in billion Rp PMSD# = Import Deflator for Services XNOS =Nominal Value of Non-oil and and Statistical Discrepancy Non-gas Export in billion Rp PX = Export Deflator XNOS$ =Nominal Value of Non-oil and PXGAS# = Price Index of Gas Export in Non-gas Exports in million US $ US $ XNOSR = Real Non-oil and Non-gas PXNOS# =Price Index of Non-oil and Exports in billion Rp Non-gas Exports in US $ XOIL = Nominal Value of Crude Oil
372 S. KURIBAYASHI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia
Export in billion Rp values of specific exogenous variables are XOIL$ = Nominal Value of Crude Oil changed by a fixed amount and the model Export in million US $ is solved for endogenous variables. We XOILR = Real Crude Oil Export in call this set of values of all endogenous billion Rp variables a "disturbed solution." Then. XR = Real Total Export the impacts can be assessed by comparing the disturbed solution with the control solution, the differences between them IV Multiplier Analysis usually being taken. One of the primary advantages of mak With respect to real government invest ing use of econometric models for econo ment (IGR) and real government consump mic planning is to be able to evaluate tion (CGR) , each of them was increased policy effects quantitatively. The short by 100 billion rupiahs every year in each and medium-term impacts of policy varia simulation. This is called "sustained bles on target variables are analysed with increase" hereafter. The differences Core Model-81 in this section. between the disturbed solution and the To measure the impacts of changes in control solution for real GDP. total real an exogenous variable on endogenous demand and balances of international trade variables. all endogenous variables are first of goods and services are shown in Table solved over the pre-assigned period with 5. Multipliers can be obtained by divid given values of all exogenous variables. ing the values of each endogenous variable This set of estimates of all endogenous by the changes in real govenment invest variables is usually called a "control solu ment or consumption, 100. To take real tion" or "standard solution." Second, the GDP for example, the impact multiplier
Table 5 IGR and CGR Multipliers (unit: billion rupiahs)
1983 1984 1985 1986 1987 1988 Real GDP IGR 62.5 76.4 70.7 61. 5 54.6 50.2 CGR 81. 6 104.4 100.7 91. 3 84.4 80.5 Total Real IGR 114.8 147.0 141.8 127.9 117.2 111.4 Demand CGR 136.6 183.5 182.4 168.2 157.0 151.4 X-M IGR -297.9 -422.2 -466.6 -477.6 -492.2 -521. 7 (-0.331) (-0.435) (-0.458) (-0.447) (-0.438) (-0.442) CGR -171. 2 -316.3 -356.1 -349.7 -341.8 -349.3 (-0.190) (-0.326) (-0.350) (-0.327) (-0.304) (-0.296) Notes: 1. These figures are based on sustained increase of 100 billion rupiahs in real government investment and consumption respectively. 2. The figures in parentheses are measured in billions of U. S. dollars. 3. X =exports of goods and services at current prices. M= imports of goods and services at current prices.
373 Jl(M7'~"1'iJm 24~4-8"
Table 8 QOIL, PXOIL, and RFEX Multipliers'
1984 1985 1986 1987 1988 1989 Growth Rate of QOIL , 0.08 0.11 0.10 0.08 0.07 0.05 Real GDP PXOIL 0.04 0.05 0.05 0.03 0.02 0.01 (%) RFEX 0.44 0.56 0.50 0.39 0.30 Total Tax QOIL 80.0 275.5 530.0 859.4 1276.9 1783.1 (billion PXOIL 84.4 270.3 494.0 766.8 1074.3 1420.5 rupiahs) RFEX 595.8 2440.6 4792.2 7620.6 10933.0 X·M QOIL 0.120 0.256 0.413 0.597 0.813 1. 043 (billions of PXOIL 0.133 0.265 0.415 0.573 0.726 0.885 U.S. dollars) RFEX 0.744 1.593 2.668 4.064 5.863 Notes: 1. Sustained one-percent increase for QOIL and PXOIL. 2. Sustained five-percent increase for RFEX.
is 0.625, and the medium-term multiplier crude oil (PXOIL) and foreign exchange reaches its highest value, O. 764, in the rate (RFEX), and are shown in Table 6. second year and declines in subsequent Comparison between QOIL and PXOIL years. In developed countries such as the shows that the sustained one-percent in United States and Japan, the impact crease in crude oil production has a larger multiplier is usually around 1. 3 and the effect than the same increase in the ex medium-term multiplier peaks at around port price of crude oil. Simply dividing 2. O. These low multipliers for real govern the results for RFEX by five tells us that ment investment in the Indonesian economy a sustained one-percent devaluation of are ascribed to high leakage through direct foreign exchange has the same effects on alld indirect imports, as mentioned above the growth rate of real GDP as the same concerning import functions. According increase in crude oil production, but much to an econometric model for the Republic larger effect on tax revenue and trade of Korea constructed by the Economic balance. Research Institute~: Economic Planning Agency, Japan, government investment V Concluding Remarks multipliers are also less than 1. O. This may be a common feature of developing As mentioned in Repelita IV, macroecono countries. In comparison with IGR multi mic models were employed in. order to pliers, CGR multipliers, which are slightly obtain a better consistency in the planned larger than 1. 0 at the peak, ,are larger. macroeconomic aggregates. Numerous Instead of a sustained, fixed amount of policy simulations. were conducted with increase in an exogenous variable, the im the macroeconometric model before a final pacts of a sustained percentage increase conclusion was reached. In our experience were estimated for the quantity of crude with Indonesian development planning, oil production (QOIL), export price of macroeconometric models can playa central
374 s. KURIBAYASHI: A Medium-term Macroeconometric Model for Economic Planning in Indonesia role and give fresh and deep insights into be closely related to the energy sub the problems and issues in the future and model which has not yet been com policy measuses for dealing with them. pleted. One advantage of employing macro (3) If available, data on imports of invest econometric models for economic plan ment goods which are directly related ning is that the actual course of the to government investment should be economy can easily be compared quanti compiled and used in the core model. tatively with the planned one, and the (4) It is desirable that non-oil exports planned policy measures can be revised should be disaggregated at least into if necessary, while implementing the plan primary commodities, manufactured year by year. This means that a rolling goods, and services. plan can be introduced. (5) Reliable data on labor force, employ For that purpose, the macroeconometric ment, and population have to be com models have to be re-estimated with ex piled, because they belong to the most tended data and revised every year if important economic indicators in the necessary. The core model and the fiscal Indonesian economy. submodel have so far been re-estimated (6) Indonesian national accounting data three times. Fairly stable parameters were do not include national disposable in estimated for structural equations, although come and its appropriation accounts a one-year extension of the data has a and income and outlay accounts by significant influence because of the small institutional sectors. Although this ness of the sample size. shortcoming is common in developing Needless to say, there exist several countries, high priority should be shortcomings in the core model, which given to their compilation. If they are are mainly attributable to lack of data. provided, our models will be markedly These shortcomings have to be remedied improved. and the core model revised in the follow (7) Our macro models will be improved ing directions. and refined by comparing them with (1) Price equations have to be revised, macro models of other developing introducing wage rate and govern countries. In other words, some study ment-regulated price index into them. and research will be needed on con In particular, the investment deflator structing the same kind of macro function needs to be re-specified models for other developing countries. without money supply in the explana tory variables. References (2) It is desirable that the quantity of Ezaki, MitSllO. 1982. A Monetary Submodel of refined oil for domestic consumption the Bappenas Model. National Development Planning Agency (Bappenas) , Indonesian should be disaggregated into com Government. (Mimeographed) ponents such as kerosene. This will Indonesia, Bappenas. 1982. Model Makroekonometri
375 Inti Indonesia. Bappenas, Indonesian ---. 1983b. Multiplier Analysis by the Macro Government. (Mimeographed) Econometric Model. Bappenas, Indonesian Indonesia. Bappenas; and Ministry of Finance. Government. (Mimeographed) 1982. An Econometric Analysis oj Indonesian Odano, Sumimaru. 1983a. Re-estimation of the Fiscal Sector--A Tentative Report--. Bap Ezaki Model ojthe Indonesian Monetary Sector. penas, Indonesian Government. (Mimeo Bappenas. Indonesian Government. (Mimeo graphed) graphed) Kobayashi, Kazumi. 1982. Indonesia's Economic ---,. 1983b. Indonesian Balance oj Payments. Development and Bappenas Macroeconometric Bappenas. Indonesian Government. (Mimeo Model, 1969-1980. Bappenas, Indonesian graphed) Government. (Mimeographed) Oshika. Takao. 1984. Revised Core Macro Econo Kuribaya,shi. Sei. 1982. Revised Core Macro Econo metric Model Re-estimated. Bappenas, Indo metric Model. Bappenas. Indonesian Govern nesian Government. (Mimeographed) ment. (Mimeographed) The Technical Team. 1982. Bappenas Core Macro ---. 1983a. Revised Core Macro Econometric econometrtC Model. Bappenas, Indonesian Model Re-estimated. Bappenas, Indonesian Government. (Mimeographed) Government. (Mimeographed)
3'18