Peranika 10(3), 375 - 380(1987)

COMMUNICATION IAn Inter-industry Analysis of the West Malaysian Economy

ABDUL AZIZ ABDUL RAHMANDepartment of Economics,

Faculty of Economics and Management,Universiti Pertanian Malaysia,

43400 Serdang, Selangor Darul Ehsan, Malaysia

ABSTRAK

Kertas ini menganalisis struktur saling berhubungan ekonomi Malaysia Barat berasaskanjadual antara industri bagi tahun 1982. Empat jenis saling bergantungan telah diperiksa, iaitu, syerperindustrian dalam komponen-komponen permintaan akhir dan input primer, keperluan inputserta urusniaga perantaraan industri-industri. Kajian ini menunjukkan sebahagian besar penggunaanswasta dalam ekonomi merangkumi per belanjaan ke atas makanan dan perkhidmatan. Struktur pern -bentukan modal tetap dikuasai oleh industri-industri pembinaan dan kenderaan bermotor, manakalagubahan eksport pula dikuasai oleh industri industri keluaran petroleum, pemprosesan getah danminyak serta lemak. Pada umumnya, industri-industri perkilangan bukan-asastani merupakan pem-beli utama input yang diimport manakala industri-industri pertanian dan perkhidmatan merupakanpengguna utama input primer. Analisis struktur urusniaga perantaraan menunjukkan sedikit sahajaindustri yang mempunyai rantaian teknologi yang meluas dengan bahagian-bahagian ekonomi yanglain. Majoriti industri-industri yang berantaian tinggi, merangkumi industri yang terlibat dalampemprosesan input yang diimport untuk pasaran tempatan atau eksport seperti baja-baja kimia,keluaran-keluaran plastik dan keluaran-keluaran logam.

ABSTRACT

This paper analyses the structural interrelationship of the West Malaysian economy based onthe 1982 inter-industry table. Four types of direct interdependence are examined, namely, industrialshare of final demand and primary input components, input requirement and intermediate transac-tions of industries. The study indicates that a substantial proportion of private consumption in theeconomy comprises expenditure on food and services. The fixed capital formation structure isdominated by the construction and motor vehicle industries while the export mix is dominatedby the petroleum products, rubber processing and oils and fats industries. In general, the non-agro-based manufacturing industries are the major purchasers of imported inputs whereas the agriculturaland service industries are the major users of primary inputs. The analysis of structure of intermediatetransactions indicates that there are only a few industries that possess extensive technological linkageswith the rest of the economy. The majority of high-linkage industries comprise those that are involvedin the processing of imported inputs for domestic or export markets such as chemical fertilizers,plastic products and other metal products.

INTRODUCTION alternative analytical techniques and there isThe development of inter-industry models com- now a wide variety of inter-industry models inmenced in 1936 with Leontiefs empirical model use. The foremost advantage of the inter -of the United States economy. Subsequent accu- industry approach of analysis is its stress on in-mulation of statistical material in the inter- dustrial relationships and interdependence,industry framework led to the development of showing empirically how "everything depends

ABDUL AZIZ ABDUL RAHMAN

upon everything else."This paper analyses the inter-industry

relations in the West Malaysian Economy, withspecific reference to direct structural inter-dependence.2 It will serve as a useful source ofinformation on economic relations in WestMalaysia and provides some guidelines fornational development planning.

INTER-INDUSTRY TABLEAn inter-industry table is a two-way transactionsmatrix of the national economy. The entries inthe table are all in money values. Within thetable, each industry has a row and a column att-ributed to it. The size of the table may varydepending upon the degree of detail incor-porated in it.

The rows of the table show the disposal ofthe output of each industry: sales of commoditiesto other industries to be utilised in the currentproductive process and sales to final demand,comprising government and private consump-tion, stocks, capital formation and exports.

The columns show the sources of the inputsutilised by each industry to produce its output:intermediate inputs purchased from domesticindustries and primary inputs. The primaryinputs comprise value added (the sum total ofwage and non-wage factor payments) and im-ported inputs. As the input column includesprofits, total inputs used in the industry alwaysequal the value of total output.

The value of all the purchases made by anindustry for each dollar's worth of output may bemade by dividing each input value by the corres-ponding total output. The resulting values areknown as input-output coefficients, and thetable of input-output coefficients is termed thetechnology matrix.

DATAThe primary source of data for this study is the1970 input-output table for West Malaysia pub-lished by the Department of Statistics (1975).Several adjustments have been made to the basictable, including the disaggregation of agricul-tural sectors into constituent crop industries andthe estimation of land development works by theagricultural contractor services subsector. Thistable is then updated to 1982 by the use of the bi-proportional rating, or RAS3, method.

The RAS method proportionally adjust rowand column totals to agree with some predetermined values. The proportions are expressed asthe vectors "r" and "s'\ being termed row andcolumn multipliers respectively. The predeter-mined values are those of final demands,primary inputs, sectoral outputs and, therefrom,the values of the row and column total vectors.The updated coefficient .matrix, A ^ is related tothe base year matrix, A Q, by the relation:

The estimates of final demands, primaryinputs and sectoral outputs for 1982 werederived from the National Accounts worksheetsof the Department of Statistics. To conform tothe industrial classification scheme adopted inthis study, some adjustments were necessary.Also, for a number of industries, indices ofvolume of output and prices were applied to theindustrial output for 1970 in order to derive theoutput levels for 1982.

RESULTSFour types of direct relations will be examined:industrial share of final demand componentsand primary inputs, direct input requirementsand intermediate transactions of industries.

This study covers West (or Peninsular) Malaysia only. This is largely because sufficient inter-industry data for analysingstructural interdependence are not available for East Malaysia (that is Sabah and Sarawak).2 For an explicit exposition of the direct, indirect and total interdependence see, for example, Miernyk (1957) and Chenery andClark (1959).3..

RAS" is a shorthand way of describing a technique which involves prorating the row and column coefficients of an inter-industry table to equal to some known marginal totals. The RAS is not the only method of updating the inter-industry matrix;for a discussion of alternative techniques, see Tilanus (1966).

4 For details of the theoretical scheme and analytical framework underlying the RAS method, see, for example, Bacharach(1970) and Stone, Bates and Bacharach (1974).

376 PERTANIKA VOL. 10 NO. 3, 1987

AN INTER-INDUSTRY ANALYSIS OF THE V/EST MALAYSIAN ECONOMY

Industrial Share of Final DemandApproximately 51% of government consump-tion comprised expenditure on public adminis-tration and defence. In addition, about 38%and 11 % of government consumption wereexpended on social services and other commu-nity services respectively.

A substantial proportion of private con-sumption expenditure was incurred on servicescomprising trade, transportation and commu-nication, house rentals and personal services.The expenditure on trade, particularly on hoteland restaurant services, accounted for 19% oftotal private consumption while that on houserentals accounted for 13%. Private consumptionof transportation and personal services weremuch smaller, in the order of 7% and 5% res-pectively of the total expenditure.

The next largest expenditure by consumersin 1982 was on food. The livestock, meat anddairy products, grain mills and other foodindustries each accounted for over 3% of totalprivate consumption. The bakeries and otherprocessed food industries also recorded a relativeshare of total private consumption of over 1 %each.

The fixed capital formation structure wasevidently dominated by the constructionindustry. It contributed 64% to the total value ofcapital goods produced in the economy. Themotor vehicles and non-electric machineryindustries, with relative shares of 8.2% and1.1 % respectively, were two other industries thatcontributed significantly to gross fixed capitalformation.

With respect of fixed capital formation inagriculture, the smallholder sector's contribu-tion was slightly larger than that of the estatesector. For instance, in the case of rubber, thesmallholder sector had a share of 2.6% while theestate sector had a share of only 1.4%. Therelatively low level of capital formation in therubber estate sector was primarily due to thedecline in the rate of new and re-planting in thesector coupled with the conversion of numerousrubber estates to oil palm cultivation.

By 1982 the petroleum products industryhad overtaken rubber processing and oils andfats to become the most important export earnerfor West Malaysia. Petroleum products account-

ed for 22% of gross export earnings comparedwith rubber and oils and fats (chiefly palm oil)whose relative shares were 19% and 18% respec-tively. The basic metals industry, another tradi-tional export industry, contributed only 12% tothe gross export receipts.

Industrial Share of Primary InputsIn general, the non-agrobased manufacturingindustries were the major users of importedinputs. Of the top ten industries in terms ofrelative share of imports, four, namely, motorvehicles, petroleum products, basic metals andother metal products, belonged to the non-agro-based manufacturing category. By and large,the agricultural industries were insignificantusers of imported inputs except for padi andrubber estates.

The government sectors of public adminis-tration and social services together accounted foralmost one third of the total wages generated inthe economy. In comparison, the proportionatecontributions to total labour payments by agri-cultural industries were insignificant except forrubber estates, rubber smallholdings and live-stock . 11 should be noted that despite thediminishing importance of the rubber sector inthe export structure, it still remained significantin terms of capacity to employ labour.

The agricultural industries, however,proved to be the major generators of non-wagefactor payments, (other value added), in theeconomy. Of the top ten industries in terms ofvalue added generation four — rubber small-holdings, padi, oil palm estates and livestock —consisted of agricultural industries. Much of thenon-wage value added generated in these in-dustries comprised depreciation charges asso-ciated with plantation and land development.Other industries which were important from theviewpoint of value added generation includetrade, dwellings and transport and communica-tion.

Direct Input RequirementsThe agricultural industries, by definitionprimary industries, had comparatively largeaggregative primary input requirements. Foreach industry the wage and non-wage factorrequirements together constituted four-fifths ormore of total primary input requirements. Like-

PERTANIKA VOL. 10 NO. 3, 1987 377

ABDUL AZIZ ABDUL RAHMAN

wise, the service industries were also moredependent on primary inputs compared to inter-mediate, or non-primary, inputs. Except forpersonal services and other services, the primaryinput requirements of the service industriesconstituted more than three-fifths of unit valueof production.

Employing the relative size of wage coeffi-cient as a proxy for labour intensity, it was foundthat only four manufacturing industries couldcategorically be classified as being relativelylabour intensive. These consisted of sawmills,furniture, paper products and industrialchemicals. In these industries, the wage coeffi-cients were larger than the non-wage valueadded cofficients. Other labour intensive indus-tries include education, personal services andpublic administration.

The comparatively high intermediate inputcoefficients of manufacturing industries such asmeat and dairy products and canning of fruitscharacterised one activity typical of a developingeconomy, namely, primary processing of localraw materials for export. These industriesrequired a high level of domestic (intermediate)compared to imported inputs. Concomitantly,they showed a strong orientation for export asreflected by their prominent shares in the exportmix.

On the other hand, the relatively highimport relative to intermediate input coefficientsrecorded by industries like wearing apparel andplastic products characterised another activityprevalent in a developing economy, namely,processing of imported inputs for domestic orexport market. For these industries the importrequirements constituted one-third or more ofunit value of production. Also, in the case ofindustries such as animal feed and other metalproducts, the low degree of processing on im-ported inputs was reflected to an extent by theirlow wage and non-wage factor input coefficients.

The average merchandise (intermediateplus import) input coefficient of manufacturingindustries in 1982 was found to be 0.62. In addi-tion, the average intermediate input coefficient(0.41) of manufacturing industries was also

greater than the imported input coefficient(0.23). In comparison, the average merchandiseinput and intermediate input coefficients formanufacturing industries in 19705 were signifi-cantly greater at 0.70 and 0.48 respectively.These observations, inter alia, indicate that theindustries involved in the processing of domesticresources for exports had expanded less rapidlycompared with other manufacturing industriesover 1970-1982.

Intermediate TransactionsThe structural interdependence between in-dustries may also be examined by analysing thecontribution of each industry to total supply anddemand.

These two measures of direct inter-dependence were computed for each industry.The individual industries were in turn ranked,following Chenery and Clark (1959), in a four-way classification of industries. The ranking wasbased on whether the industrial shares in totalsupply and demand are above/below the corres-ponding economy-wide averages.

Thirty-four out of a total number of sixtyindustries sold two-thirds or more of their out-puts directly to final demand, mainly to privateconsumption and export sectors. Further, twentyof these industries comprised manufacturingindustries.

The significance of agricultural industriesas major suppliers of inputs for agrobased manu-facturing industries was reflected by the fact thatthey sold over two-thirds of their outputs tointermediate demand.

On the other hand, the service industries,with an average of 20%, sold proportionatelyless of their products to other industries. Like-wise, the manufacturing industries as a wholerecorded a relatively low average intermediatedemand proportion of 27%. This indicated thatin general the manufacturing activity was stillstrongly oriented towards the final markets,especially for consumer and non-durable goods.

The industries included under Category IIand Category III were characterised by high(direct) backward linkage values. Accordingly,

These coefficients were computed from the 1970 input-output table

378 PERTANIKA VOL. 10 N o . 3, 1987

AN INTER INDUSTRY ANALYSIS OF THE WEST MALAYSIAN ECONOMY

they depended heavily on other domestic in-dustries for their inputs. Changes in their level ofactivity would concomitantly generate greaterrepercussions on the economy than those includ-ed in either Category I or Category IV. CategoryII industries, which also had high (direct) for-ward linkage values, comprised predominantlynon-agrobased manufacturing industries. Onthe other hand, Category III industries, whichwere characterised by low forward linkageindices, consisted largely of agrobased manu-facturing industries.

Taking an overall measure of direct inter-dependence, the industries in Category II, withcomparatively high backward and forwardindices, would possess the most extensive struc-tural linkages in the economy. A total of nineindustries fell under this category, seven of whichwere non-agrobased manufacturing industries.Almost invariably, the latter industries com-prised those that processed imported inputs fordomestic or export market such as chemicalfertilisers, paints and plastic products.

At the other extreme, industries in CategoryIV, with low backward and forward linkagevalues, would be relatively independent of otherindustries for both their inputs and outputs. Alarge number of service industries includingtrade, education and hearth belonged to thisgroup.

CONCLUSIONSThe application of inter-industry analyses forexamining the structural interrelationships ofthe national economy has been widely described.A useful feature of an inter-industry model isthat it measures, explains and compares thedegree of interdependence among the varioussectors of the economy.

This paper has analysed the inter-industryrelations in the West Malaysian economy. Fourtypes of direct interdependence were examined:industrial share of total final demand and totalprimary inputs, input requirements and inter-mediate transactions of industries. The basicdata for this study were provided by the 1982input-output table for West Malaysia.

The study indicated that a substantial pro-portion of private consumption in 1982 com-prised expenditure on services and food. These

two items together accounted for approximatelyone-third of total private consumption expen-diture. The government consumption expen-diture, on the other hand, was confined only tothree sectors — public administration, socialservices and other services.

The fixed capital formation structure of theeconomy was dominated by the construction andmotor vehicle industries whose respective shareswere 64% and 8%. The level of capital for-mation in plantation and land developmentwas predominant in rubber and oil palm indus-tries, with the smallholder sector generallyrecording a greater level of investment comparedto the estate sector. In terms of exports, it wasfound that by 1982, the petroleum productsindustry had surpassed the traditional exportindustries of rubber processing, oils and fats(palm oil) and basic metals (tin) industries tobecome the most important export earner forWest Malaysia.

Generally, the non-agrobased manufactur-ing industries were the major purchasers ofimported inputs. In contrast, the agriculturalindustries were insignificant users of importedinputs except for padi and rubber estates. Theservice industries together contributed almostone-third to the total wages generated in theeconomy while the agricultural industriestogether accounted for approximately one-fifthof the gross non-wage factor payments.

The agricultural and service industriesgenerally utilised high levels of primary com-pared with non-primary, or intermediate,inputs. In comparison, a number of industrieslike meat and dairy products and canning offruits utilised proportionately more of inter-mediate than primary inputs. Almost invariably,these industries comprised resource-based manu-facturing industries whose production wereoriented towards exports. In the case of assemblyindustries like wearing apparel and plastic pro-ducts the imported input requirements weresubstantially greater than the intermediate inputrequirements.

The analysis of structure of intermediatetransactions indicated that there were only a fewindustries that had extensive technologicallinkages with the rest of the economy. Themajority of these strategic industries were

PERTANIKA VOL. 10 NO. 3, 1987 579

ABDUL AZIZ ABDUL RAHMAN

involved in the processing of imported inputs fordomestic or export market such as chemicalfertilizers, plastic products and other metalproducts. Accordingly, any changes in their levelof productive activity would generate extensiverepercussions on the economy.

Some comments are necessary concerningthe shortcomings associated with this study.Obviously, the usefulness and validity of theresults presented above are limited by therealism of the inter-industry model employed.For instance, the model is based on currentflows, and it assumes fixed technological coeffi-cients. This static model is open to questioningbecause it ignores the possibility of factor substi-tution. It also ignores several other difficultiesinvolved in adopting a static inter-industryframework as a basis for analysing problems ofstructural interdependence. From a differentviewpoint, the measures of industrial inter-relationships are also bound by the particularindustrial classification scheme adopted in thebasic inter-industry table and price changes.

REFERENCES

BACHARACH, M. (1970): "Biproportional Matrices andInput-Output Change.'1 Cambridge Univ. Press.

CHENERY, H.B., P.G. CLARK. (1959): "Inter-industryEconomics." New York: John Wiley.

LEONTIEF, W.W. (1936): "Quantitative input-outputrelations in the economic system of the UnitedStates" Review ofEcons. and Stats, 18: 105 - 125.

MALAYSIA. (1975): Department of Statistics. Input-output Tables for Peninsular Malaysia 1970.

MlERNYK, W.H. (1957): "The Elements of Input-out-put Analysis". New York: Random House.

STONE, R., R. BATES, M. BACHARACH. (1974): Cam-bridge Univ., Department of Applied Economics."A Programme for Growth, No. 3. Input-outputRelationships 1954-1966" London: Chapmanand Hall.

TILANUS, C.B. (1966): "Input-output Experiments.The Netherlands 1948-1961" Rotterdam Univ.Press, Econ. Series, Vol. 5.

(Received 15 December, 1986)

380 PERT ANIKA VOL. 10 NO. 3, 1987