Issues in NI of Pakistan

7/30/2019 Issues in NI of Pakistan

1/15

3

N AT IO N A L I N C O M E A C C O U N T I N G I N PA K I STA N

by Salimullah Khan1I N Pakistan, the Central Statistical Office is charged with theresponsibility of preparing national income estimates of thecountry. These estimates are required for the measurement ofthe performance of the economy during the Plan periods and forthe evaluation of the effects of specific policy measures. The workwhich has so far gone into the production of reliable incomeestimates and other related sub-totals has already brought tolight a num ber of deficiencies in economic and financial statistics.Efforts are now under way to fill in the gaps in information undera system of priorities which ensure a most economic andjudicious use of the available resources. The availability of therequired da ta would enable us to improve the present estimates.This paper presents some of the difficulties which we havefaced in compiling the national income estimates of Pakistan.

U p to now, three sets of national income estimates have been

prepared by the Central Statistical Office in Pakistan. The k s tof these was prepared in 1950 for th e year 1948-9 in currentprices an d utilized the concept of net dom estic product at factorcost.2 Subsequently, with the availability of more data, estimatesfor 1949-50 t o 1952-3 in terms of current prices were com putedand published in the February 1955 issue of theStatisticalBulletin. These were then revised and published in the 1955Statistical Year Book together with 1953-4 national incomeestimates also in current prices (see TableI). Estimates for sub-sequent years at current prices have not yet been computed.However, in accordance with the recommendations of the UnitedNations,% nd the emphasis laid in the First Five-Year Plan* onthe measurement of economic growthin real terms, a third set of

1 Central Statistical Office, Pakistan. The author is grateful to Dr. E. F.Szczepanik of the Harvard Advisory Group, Planning Commission, for hiscomments and suggestlous in preparing t h ~ s aper

V e e S!atis!icul B~cllenn,Vo. I , No. 3 of May 1952. Since, in order to fill in themany gaps in statistical data, a number of expedients and assumpt~ons wereemployed, these est~mates are rough calculations of the goods and servicesproduced m the country, and are no longer used by the Government of Pakistan.

Un~ted Natlons, A System of National Accormts m ~dSw ppor!ingTobles, NewYork, 1953, p. VII.

The hm! Ftve Year Plan, Karachi, 1958, p. 125.47


2/15

48 M I D D L E E A S T E R N STUDIES

national income estimates has been developed in terms ofconstant prices, which have been continued to date (see Table 11).This new series at constant prices is not comparable with the

previous series a t current prices and is the only table publisheda t present.

Conceptual Dz$culties and Limitationsof Available DataIn line with other underdeveloped countries, Pakistan shares

the difficulty in applying some of the concepts and methods ofnational accounting used in the industrialized economies.Another major difficulty consists in the complete absence ofsome data and unreliability of the available data.

A distinct feature of the Pakistan economy is that it lacks aclear-cut diierentiation in economic functions inasmuch as alarge portion of its economy consists of household enterprisesperforming functions which would normally fall underdifferentindustrial categories. Sizeable groups, especially among agricul-tura l producers, pursue other occupations, often in urban areas,outside the main industry in which they are supposed to beengaged. Hence, the customary classification of national incomeby industrial origin, which is based on the concept of incomeoriginating in each single branch of economic activity, cannot beapplied except possibly as a broad and rough approximation to aclassification of the population into di ie rent groups whose mainincome is derived from one single industry.

The problem of measurability is further complicated inPakistan by the fact tha t a considerable portion of output iseither consumed by the producer himself or bartered away for

other commodities and services. It simply does not enter themarket for exchange into money through the usual commercialchannels. 'This raises a problem of including the whole of suchproduct in the national income estimate and of evaluating it interms of money.'l Even in industrialized countries, where theproblem is relatively insignificant, statisticians have often com-plained of this dficu lty,2 which presents a formidable ob for theincome estimator. Moreover, the resulting under-valuation of

national income imposes a serious limitation on international1 V.K.R.V. Rao, 'Some Reflections on the Comparability of Real National

Income of Industralized and Underdeveloped Countries', in Income and Wealth,I.A.R.I.W., 1953, p. 181.

See 'Treatment of Government Economic Activity in the National Accounts'by Ingvar Ohlson in: Income and Wealflr, Series 111, I.A.R.I.W., 1953, p. 224.


3/15

S A L I M U L L A H K H A N 49

comparisons, particularly between the underdeveloped and theindustrialized countries.

Many producers have only a very vague idea of the quantity

and value of their output. Book-keeping is very rare among pro-ducers and consumers, the vast majority of whom are illiterate.Even in western countries, most households and m any firms dono t keep any accounts. For households, family budget enquiriescan be conducted but for business enterprises case study maybe the only practical solution. In absence of these, only guess-work remains.

Apart from the conceptual difficulties mentioned above, thereis a much larger problem of the non-availability of compre-hensive statistical data regarding almost all sectors of theeconomy for the estimation of national income and relatedaccounts in Pakistan. Com paratively few economic data arecollected according to the requirements and specZications of thenational income estimator; and the gaps in technical informationand statistical data are particularly glaring, especially in the caseof complex economic structures and activities. Where theavailable information is conceptually adequate, it may bein-

accurate due to the limitations of the collection process. Thereare errors in observation, reporting and recording. Thus, forexample, price and cost data, especially for agricultureandrelated activities, are very unsatisfactory; manufacturing dataare available only for large-scale industry; there is no adequateinformation on the structure of costs, o r on consumer expendi-tures and savings; some of the items in the international balanceof payments are rough estimates; there are no satisfactory dataon distribution of income or capital formation.

Whereas the general lack of statistical da ta may continue forsome time, it is also true that there havebeen some improve-ments. The population censuses of 1951 and 1961, censuses ofmanufacturing industries and several family budget enquirieshave provided much of the needed information. There are sixmain sources for national income estimation inPakistan:

(i) Data on agricultural production and prices.(

Data on large-scale manufacturing industry, includingthe volume of output and cost structure.(iii) Some very inadequate information on the working popu-

lation, which permits derivations, in some cases as


4/15

50 M I D D L E E A S T E R N S T U D I E S

residuals, of persons engaged in cottage and small-scaleindustries, professional and related services, domesticservants, etc.

(iv) Inform ation on the revenue and expenditure of govern-ment and Government enterprises.(v) Data on exports, imports and the international balance

of payments. However, even these data are not completeand are often inaccurate.

(vi) Information on income and expenditure, rural indebted-ness and cottage industries in rural areas made availableby the National Family Expenditure Survey 1955-6 fortwenty-one urban centres, and by the Family ExpenditureSurvey 196@1 for rural areas.

The two parts of Pakistan, namely, East and West Pakistan,have somewhat different economies with different habits of con-sumption and somewhat different production patterns. In-adequacy of data therefore cannot easily be overcome byextending data for one region to another. This necessitates thecompilation of da ta on a regional basis.

The Present Estimates

Admittedly the existing methods of national income account-ing in Pakistan require substantial improvement. The m ain tablepublished by the Central Statistical Office is entitled NationalIncome of Pakistan (see Table 11), and it gives the distributionof national product by industrial origin at constant (1949-50 to1952-3) prices. This table is unsatisfactory in many respects.For instance, the classification of sectors given in this table doesnot conform to the U.N . procedure. It is an obsolete classifica-tion introduced some twelve years ago and needs change andfurther elaboration of its contents.

The income concept a t present used in Pakistan is that of netnational product a t factor cost. It equals the net aggregate outputof all the goods and services producedin various sectors of theeconomy during the year. It has been valued exclusive ofindirect taxes, and subsequently adjusted for net factor income

payments. The estimates are further adjusted for the effects ofthe terms of trade and indirect taxes on exports.About 65 per cent of the total national income has been

estimated by means of the product approach, and the rest withthe aid of the income method. Except for some minor improve-


5/15

'aleredas 01 i I n 3 W s! qqqm 'saqrd asaqlO?II! sla8 lsoa il od so el lj o iuama1a ue slayrem {eao {val ean aql uro~jpal3anoa resaa!rd alesaloqm aau!g .saSeiua3rad u!e l~ aa q Bu!leyap ~a l j easn uaaq aneq saqrda[esa[oqm 'sasea q m s u~ 'alqe[!enE IO U a l e sa>!Jd isahlaq 'sdors Kueur I O ~ ,

'lsanieq jo am!laqi i e slayrsm p rn r aq l u! sa3pd aIssa[oqm aq l LfIEnsn a l e s aq ld ?sanreH,

s! uelspled u! e3ap a3!xd ay$ jo suogm~ur!~noyzas aqr jo a u o'3)a 'sBmppqUIIVJ

'quauraldury le~nqn9y~Be0 uogepa~dappue i!edaJ 'a~ueua$-u!em ~ o jua3 lads pue l a % e l s e ~OJ sdo.13 xaqtoJOJ aua3 la d 1pue s d o n leala3 xoj $ua3 ad& :spaasJOJ 3ua3 lad ~ j oog3npaplal je pa leq1sa uaaq seq pappe anleh lau ayJ .anIeA ssoBu!e~qo03 naX q3ea u! uo!pnpord apyM-$qx$s!p Xq pagdylnmale , s a ~ p d sahley (&-256103 05-6~61) agelahe apIM-l3!1$SypaqL r.sa3pd ts aueq pue e p p jo lndlno le3ysXyd uo paseqale u o y n p o ~ ddo^:, Lq pappe a n p o saleqysa luasaxdlv

.JalcJexeq3 u! a~!vaCqns 1aql-a a le p u tuoyleqrsa jo spoqam3gzluays no paseq lou ale e p p uo!$Dnpoxd .paho~dmy q 01 aheyuo!ynpo~d do.13 jo s a l e q s a ayl 'a1qel!ahe a le sax@ a2ea13eaiem33e L~JFJpue sls$a ura~sKs uyp103a~poo%e qBno q~m

sa'0.12 (!)a.inzlnq.8~.v

.Molaq passnmpXpapq ale sura~qo~dsaq& .sa3ud JuehaIaI aqt pue lndmo joamnIoa ayz uo elep alqeIreAe Lpuann3 pue a w n m e Xtqeuosaa~jo y3e1 aql moxj sasylesaleuxrrsaasaql uy ssauyeah xo[em ayJ

.Lmouo3a atp uy saBueq3 IemnpnJls

jo pue q ~ o . 1 3 pxouoJa jo a)eJ aq$jouo!le31pq lnJ4u!ueam Xueany8 sa le q)s a q3ns Java& 'alojaxaq? '~nj$quopT $1 $ua3 ladOE01 oz aq Lern xoJxa jo u@em ayl pne a-[qegazun L~q%!q alesa$etqlsa3ngsyxa aql 'ynsax e s v sy.murq3uaq ysaM Xlle!luassa~ouo!$e~ode~$xap n n e y8no.rql pau!e$qo l o s3gs!le~s alep-jo-$nono paseq xaq!a axe sa$euu$sa lu as a~d q$ jo ~ e yeql aJow

.sJo$er?gapa$e l~do~dde0 p ~ eqlq l ! ~a ~ u d eisno303 pampa1 Xl~uanbasqns u s sa3pd inaxmnala apem aqEM sa2eWsa 'asmo3 anpUI .sassan%Tnq Buytpouaxe salemgsa iuauodmo3 aql jo h e w .palsn[pe Xl$uanbas-qns uaaq $ou akeq q 3 1 q ~a3pd iuaIxn3 jo asn aql sanlonv ualjoslopas uyeua3 u! uo!$am!isa jo am pa3o~d qi'13ejUI 'sa3eq$saa3y.rd-lua~m3agma am ue tp aIqega1 ssat s i~a dse mos u! pueaiernme axom ou ale satem!]sa a3pd-luelsuo3 Mau ayl 'sjuam

1s N V H X HV?'IflEIIIlVSI


6/15


the lack of clear-cut definition. The definition of harvest pricesdiffers from region to region. Although it represents pricesprevailing during the harvest period, analysis has shown thatharvest prices are frequently higher than the average of whole-sale prices for the corresponding period in small market townsor primary 'mundis' (i.e. the nearest points of sales from th efarms). Thus, a standard defition of harvest prices needs to beevolved and utilized for such official statistics.

Since no firm basis is yet available for netting the agriculturalproducts, information o n the cost structure of various crop itemsneeds to be obtained through a series of case studies of carefullyselected representative farms engaged, as exclusively as possible,in single crop production. Relevant studies in other Asiancountries may provide a check.

(ii) LivestockA t present, the income originating in the livestock sector con-

stitutes slightly more than 10 per cent of the national income andis estimated on the basis of production of major livestockproducts.

Using the 1945 (pre-partition) and the abbreviated 1949 live-stock census data, the Central Department of Co-operation andMarkekng compiled province-wide figures of major livestockproducts for 1948-9 and on an all-Pakistan basis fo r1952-3. Thefigures of 1952-3 were broken down by provinces o n the basisof ratios exhibited in the 1948-9 figures. The figures foriuter-vening years were obtained by interpolation and for the sub-sequent years were either available in the aforesaid department o rrepeated from previous years.

To tal output figures are obtained by applying factors of yieldper animal. None of this information appears to be valid and theestimates should be regarded as very rough approximations asthey are based on old and out-of-date surveys. At present, inaddition to this, some data are available on the number ofanimals from the abbreviated 1955 Agricultural Census Report.More complete counts and breakdowns will be published in the1960 Agricultural Census Report. Interpolation between 1955and 1960 as well as extrapolation after 1960 will be thus possible.I t will, however, be desirable tha t an Agricultural Sample Surveyshould b e taken every five years to provide benchmarks. Quicksample surveys of consumption and sales can be used as a check


7/15


on the quantity estimates obtained from the interpolation andextrapolation of the Census data.

Monthly wholesale prices of livestock products are availablefrom the Central Departmeut of Co-operation and Marketing.In order to obtain annual average producers' prices, the whole-sale prices by commodity were deflated by ratios varying from20 to 25 per cent by kinds of livestock products, determinedthrough some partial surveys and in consultationwith govern-mental authorities. Until some scientific case studies are madeon the markup of producers' prices to wholesalers on an areasample basis it is diEcult to ascertain the validity of the presenttechnique. Since no information is available on the maintenance

of livestock and the cost of materials and depreciation of imple-ments, no attempt has so fa r been made to arrive at the net valueadded of livestock products, which constitutes a serious short-comingin the present estimates.

(iii) ForestryPresent estimates for forestry are based on the value of forest

products reported by the Ministry of Agriculture and by theChief Conservators of Forests, East and West Pakistan; butthe figures as available from the M inistry of Agriculture and theChief Conservators of Forests a re a t variance. N o informationregarding the value of products of private forests is available andtherefore the contribution of forestry is an underestimate. Aerialphotographs recently taken may help to provide information fo rprivate forestsin due course.

Auction prices are used fo r the evaluation of forest productsand do not include cost of cutting and transporting to th i nearest

railhead. To bring it in line with other a~ ricu ltura l ectors. it isnecessary to add &e value of these activitgs to the anction$ices.

(iv) FisheriesProduction figures for both marine fish and inland fisheries

are supplied by the Central and Provincial Fisheries Depart-ments. I n the estimates ofinland fisheries outpu t of East Pakistanespecially, widely diverging figures have been reported by theProvincial Fisheries Department. These divergencies betweenthe two sets of estimates need to be removed. The facts arobablvX~ ~ ~ - 2can best be estimated from surveys of household expenditureand consumption both in the rural and urban areas. Perhaps the


8/15


present National Sample Survey programme carried on by theCentral Statistical Office will produce reliable measures ina fewyears time.

For estimating the contribution of fisheries, only the retailprices were available in the case of East Pakistan. These werearbitrarily reduced to obtain prices received by the fishermen inmral and urban areas. In the case of West Pakistan, wholesaleprices were available and were reduced by varying appropriateratios for individual areas to obtain producers' prices. To obtainthe prices actually realized by the fishermen, mark-ups betweenwholesale and retail sale prices and between producers' andfishermen's prices need be estimated. At present, to obtain thenet income of this sub-sector a 5 per cent deduction has beenmade from the gross value for depreciation. Satisfactoryestimates for depreciation of boats, nets, etc., are necessary forthe evaluation of fish products.

B. MiningTill now, production and price data of all minerals, except

rock-salt, were collected by the Bureau of Mineral Resources.The statistics of rock-salt are compiled by the C entral Board ofRevenue. The official data on mining refer only to W est Pakistan.Miningproducts arenotcommerciallyexploitedinEastPakistan,hence they are not included in the official statistics. For estima-tion of the contribution of the mining sector to national income,data for major mining items only are processed. Since the com-mercial exploitation of certain minor items started late, constantpit-headlwell-head prices to be used could not be ascertained.

Hence, their contribution has not been included in the nationalincome estimates. The estimates are valid only to the extent thatthe prices are relevant. These prices must be scrutinized and, inany case, estimates for intermediate products consumed intheprocess of producingminerals need to be worked out.A study ofcost structure of mining enterprises needs to be undertaken andthe results of this study should be used for the estimation of grossand net value added in the mining sector.

C. Manufacturing IndustriesThe estimates for large-scale industries rest primarily o n the

1954 Census of M anufacturing Industries carried out in respect


9/15

SALIMULLAH KHAN 55

of factories employing twenty or more persons.1 This Censusprovided benchmark value added data2 in current prices, andhad wider coverage than the1953 or 1955 censuses.A wholesaleprice index of selected domestic manufactured goods was pre-pared to deflate the benchmark figuresin order to convert theminto constant prices. In order to provide a series in constantprices for earlier and later years, the interpolation and extra-polation of the benchmark data is carried o n with a quantumindex of industrial production. However, no satisfactory priceindex exists, no r have adequate prices been collected with whichto convert the constant-price measure to a current-price measure.

The Census of Manufacturing Industries a t present excludes

one of the major segments of Pakistan's industrial economy, i.e.small-scale and cottage industries. Although it constitutes a littlemore than 5 per cent of the national income, from the standpointof number of persons engaged, this sector is second only toagriculture. The present benchm ark figure which is based o n the1951 Population Census is little more than a guess as it has beenextrapolated by the annual increase in the population under theassumption that growth in small-scale industries has been inproportion to the increase in population. Hence reliable annualestimates for various types of small-scale or cottage industriesare badly needed to estimate their production value, profits,wage bill, etc.

D. Transport and CommunicationAt present, income from transport and communication con-

stitutes about 3 per cent of the national income. Transport o therthan railways is estimated from the product side on the basis of

assigning annual receipts to taxis, trucks, buses, lorries, victorias,t o n g a ~ , ~tc., the num ber of which is obtained from the Ministryof Com munications, as well as from scattered and partial data.The average net income per vehicle has been calculated on thebasis of inform ation from various sources after deducting inter-mediate expenses. Net income has been obtained by multiplyingaverage net income per vehicle by the number of vehicles.Country transport by animal-drawn vehicles is treated as p ar t of

No such census was held for 1956; and those held for 1955,1957, 1958 and1959-60 although an unprovement over the last censuses, contamed an unknownamount of undercwerage.

a Due allowance was made for non-reporting industries.Victorias and tongas are local horse-drawn vehicles.


10/15


agriculture and no attempt is made to estimate it independently.Valid estimates for cost of maintenance and depreciation ofvehicles in addition to the monthly wages paid to the drivers andconductors, etc., are not wholly available. The major short-coming appears to be in water transport, especially that pro-vided by water boats in East Pakistan. Except for thosemechanically propelled, no information of any kind relating toriver transport is available. The net income of other transportservices, composed of transport managers, road vehicle drivers,air transport workers, and sea and river craft workers, was there-fore calculated on the basis of the number of persons engaged insuch services as derived from the 1951 Population Census and

their assumed income for each category. Projection for back andsubsequent years was made on the basis of the rate of change ofpopulation, which, however, is not the correct measure of theircontribution. In the case of mechanized vehicles the estimatescan be improved with the help of sample surveys with specialreference to cost structure. For non-mechanical vehicles inurban areas, comprehensive sample surveys are the only possiblemethod of estimation. For water-transport the reports of theInland Water-Transport Authority in East Pakistan which pro-vide information on the number of employees, passengerscarried, boats, etc., need be utilized. Moreover, these estimatescan further be improved as and when further information ismade available from the 1961 Population Census.

The income originating in rail transport and communicationhas been estimated from the income side on the basis of wagesand salaries paid to the employees engaged in these services asderived from the budgetary reports of the Pakistan Railwaysand Post and Telegraph Administration, and these data arereliable. However, Railways and Post and Telegraph should betreated as commercial enterprises and operating surpluses ofrailways and postal and telegraph services (excludmg broad-casting) should be included in the relevant estimates, which isnot being done at present.

Virtually all other industries have been estimated from theincome side. That is, wages and other personal earnings have

been estimated for a benchmark year (normally 1951) byassigning to each industry average annual earnings per personengaged and expanding this by the number engaged accordingto the 1951 Population Census. This method places heavy


11/15


reliance on the accuracy of the occupational classification in theCensus, as well as on the estimates of their average annualearnings. The latter, while essentially rough estimates, no doubttake into account such information as has been obtained invarious surveys. This cannot be regarded as a reliable bench-mark for the trade and services sector, especially since these aredominated by very small concerns in which entrepreneurialprofit income predominates. No usable information on profitexists, data from the income-tax statistics are not satisfactorilyclassitied for this purpose and have inadequate coverage.1

E. Trade and Services

In the case of services, the extrapolation of the benchmarkmentioned above has been carried out on the estimated annualincrease in the population. The number of persons engaged inservices for 1950-1 was derived from the figures of occupationaldistribution as provided by the 1951 Population Census. Income-tax data and other sources were utilized to determine an averageannual income for each of the various categories of services.Income for this sector was computed with the help of the abovedata which were projected backward and forward for earlierand subsequent years. There is ample room for improvement.For instance, it is doubtful that the personal services have in-creased at the actual rate of growth of the population. Details ofoccupational distribution from the 1961 Census when madeavailable will provide better benchmark data. At least, to startwith, income originating in broad sub-groups like (a) education;(b) medical, public and private health services; (c) recreation andentertainments and (d) domestic services, etc., need be computed

for better understanding of the estimates.Figures on persons employed in wholesale and retail tradehave been obtained from the 1951 Census of Population. Aroughfigure of average annual income separately for each category wasdetermined with the help of income-tax returns. The total incomefor this sector (1950-1) was obtained by multiplying averageannual income by the number of persons by categories. The yearto year variation in the total income has been adjusted by thechanges in the value of production of agriculture, manufacturingand mining all in constant prices. The use of estimated changesin the commodity producing sectors for extrapolating the bench-

At present, tax receipts are about 8 per cent of the national income.h5.B.S.-C


12/15


marks for trade neglect the changes in the proportion of theproducts of agriculture that pass through trading channels, aswell as omitting changes in the volume and gross margins ofimports. For these reasons the trade and services estimates areweak and cannot be much relied upon as indicators of the levelof income or changes therein.

At present, construction, electricity, gas, water and sanitaryservices are included in the services sector. Some data areavailable for public construction, but there are practically nodata on private construction. The estimates of income forconstruction have, therefore, been computed on the basis of thenumber of persons engaged in this industry as reported in the1951 Population Census and the estimated annual income deter-mined on the basis of income-tax data and some wage surveys.The benchmark for 1950-1 has been projected backward andforward in relation to the changein population growth. This isnot a satisfactory basis for evaluating the income originating inconstruction. These do not correspond to the present dayconditions and cannot therefore be regarded as sufficientlyreliable or accurate to be nseful.l

I n the absence of any system of reporting construction da tafor ru ral areas, the Central Statistical Ofiice has recently decidedto undertake a rural construction survey. Such surveys shouldbe conducted every five years to provide necessary benchmarks.For the intervening years intercensal rates of growth should beused. Besides these, the material on construction available inother agencies needs to be examined and used, if found suitable.

F. Rental Income

Estimates of rental income are based on the number of dwell-ings as shownin the 1951 Census and estimated annual rents forrural and urban areas. In order to arrive at net rental income inurban areas 50 per cent of the gross income was allowed fordepreciation and the rest was further deflated by 10 per cent formaintenance and repairs. For rental income of rural areas, only10 per cent of the tota l rental income was deducted for main-tenance and repairs. These represent net rentals for urban andrural areas for the year 1950-1. The benchmark figures soobtained have been projected backward and forward on the

1 The sam e criticism applied t o electricity, gas, water and sanitary serviceswhich should be shown as a separate sector .


13/15


basis of the rate of change in population. These estimates do no tcorrespond to the present day conditions and, therefore, cannotbe regarded as sufficiently reliable. Municipalities levy a tax on

rental income in urban areas. With the help of these records theaggregate rent can be estimated on the basis of the tax assessed.Conducting a sample survey of rents paid and imputed in ruralan d urban areas will provide satisfactory information on owner-ship of dwellings. A t present, plans a re in hand for conductingsurveys of rural rental paid or imputed and the results whenavailable will help improve the estimates.

G. GovernmentIt has been assumed that the real outputin this sector is

measured by money wages paid. Income originating in thissector is drawn from the final accounts of the central, provincial,and local authorities' budgets. The estimates are probably thebest tha t can be made at present. Since Government budgets areclassified in a manner which requires adjustment for use innational income estimation, there is a possibility that certainitems, like transfer payments, subsidies, cap ital formation, etc.,

may not receive a proper treatment. The two major difficultiesare the lack of sufficient details in the budgets a nd the continuousinability to obtain adequate coverage of local authorities'budgets.

H. Banking and Insurance(i) Banking

Income originating in banking is derived from the financialreports of the major concerns in this field an d is computed onthe basis of wages and salaries of the employees of the StateBank and the wages and salaries and the operating surplus ofthe scheduled banks. N o imputed charges for banking servicesfrom other sectors are deducted. At present, no reporting systemexists for non-scheduled banks. Relevant ratios, however, canbe determined on the basis of information concerning wages,profits and depositsin the scheduled banks. This is being takenin hand and will be introduced soon. Furthermore, no informa-

tion exists on the volume of income arising from professionalmoney lenders who playan important role in the supply of creditin the country's rural economy. Moreover, at present, semi-governmental banking institutions which are n o t scheduled by


14/15


the State Bank such as P.I.C.I.C., Industrial Development Bank,etc., are not included in the present estimates. These should alsobe included together with the various statutory bodies for publicdevelopment, e.g. P.I.D.C., W.A.P.D.A., K.D.A., etc. At pre-sent, very little information is available on transactions concern-ing the sale of land and buildings. This calls for a study of thesubject by a research institute in order to introduce an estimatefor the real estate sub-sector.

The contribution of co-operative societies is also based on thewages and salaries paid to their employees. This informationwas obtained from the Central Co-operation and MarketingDepartment. No up-to-date information on Co-operative CreditSocieties is available at present.(ii) Insurance

At present, the contribution of insurance is also estimated onthe basis of wages and salaries paid to employees and commis-sions paid to commission agents. At present, the operatingsurplus of the insurance companies, has not been included inthe estimate which is therefore too low. The contribution ofpostal insurance at present is included in the communicationssub-sector instead of the insurance sector.I. Rest of the World

At present, this sector consists of several sub-categories:(i) factor income payments, (ii) adjustment for indirect taxes onexports, and (i) effects of terms of trade. Net investment incomeis taken from the Balance of Payments statements. For estimat-ing factor income payments at constant prices, net investment isdeflated by the adjusted import price index for the current year.But the calculation of 'net factor income payments' is limitedonly to investments; other factor income should also be included.For the adjustment of indirrect taxes on exports, average exporttaxes during the base years 1949-50 to 1952-3 have been extra-polated by changes in the physical volume of exports. The adjust-ment for indirect taxes on exports does not seem to be meaning-ful. All indirect taxes should be included in order to move fromfactor cost to market price estimates. The effect of changes in theterms of trade has been computed by deflating the value ofexports in the current year by the import price index and by theexport price index. The difference between the two deflatedvalues represents the effect of terms of trade.


15/15


Detailed official statem ents on transactions with the rest ofthe world a re available. There are large differences between th ecommodity data from the balance of payments and from the

trade statistics. Otherwise the rest of the world sector does notpresent any substantial problem. Delays in the publication of theofficial estimates often make it impossible to correctly assignthe factor income receipts or payments to the year in whichtransactions actually took place.Conclusion

It will be seen that the greatest limitation in the way of im-provement of these estimates is the paucity andunreliability of

da ta and thelack of adequate reporting systems. In order to meetthe growing needs of planners, research workers and other usersof national income data, these accounts must be prepared in acomprehensive form. Efforts are now being made to improve theestimates as far as possible. Collection of statistics is a time-consuming process and will take its own time. However, it isstrongly felt that unless the Central Statistical Organizationsin the developing countries are equipped with legal authority tocollect all types of statistics it is difficult to obtain adequate andreliable basic information. In order to achieve quick and cheaperresults in these countries, sampling techniques will have to beapplied more and more to collect the needed statistics. I t mayalso be emphasized that the national accounting units in thesecountries should serve as the clearing houses for all sorts ofstatistics. No statistical data collection programme should beplanned or undertaken without the consultation of such unitsso that the statistics collected can be used for national income

accounting as well. Moreover, com peten t personnel possessingsound knowledge of economics and statistics should be attachedto the field of national income accounting for skilful handling,processing and analysis of data. It is also most desirable thatadequate training facilities should be provided to the nationalincome workers so th at they may acquaint themselves with theprocedures and techniques of national income accounting in th eadvanced countries.

Documents

Issues in NI of Pakistan