Pakistan Institute of Development Economics, Islamabad

Impact of Workers' Remittances from the Middle East on Pakistan's Economy : Some Selected
Issues [with Comments]
Author(s): Rashid Amjad and Meekal Aziz Ahmed
Source: The Pakistan Development Review, Vol. 25, No. 4, Papers and Proceedings of the Third
Annual General Meeting of the Pakistan Society of Development Economists (August 10-12,
1986) Part II (Winter 1986), pp. 757-785
Published by: Pakistan Institute of Development Economics, Islamabad
Stable URL: http://www.jstor.org/stable/41258790 .
Accessed: 27/09/2013 01:47

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Pakistan Institute of Development Economics, Islamabad is collaborating with JSTOR to digitize, preserve and
extend access to The Pakistan Development Review.

http://www.jstor.org

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 The PakistanDevelopmentReview
Vol. XXV, No. 4 (Winter1986)

Impactof Workers'RemittancesfromtheMiddleEast
on Pakistan'sEconomy: Some SelectedIssues

Rashid Amjad*

Over the last decade, the phenomenon of overseas migrationfromPakistanto

the countriesof the Middle East and NorthAfrica1has had a farreachingimpact on
the domestic economy. Indeed, no factorhas more dramaticallyaffectedthe domes-
tic employmentsituation and the balanceof-payments position as the outflow of
contractworkersand inflowof workers'remittancesfromthose countries.According
to the Sixth Plan, as much as one-thirdof the increasein the labour forceduringthe
years 1978-83, i.e. the Fifth Plan period, was absorbed by migrationto the Middle
East [14, p. 499]. At its peak in 1982-83, official flow of remittancesfrom the
Middle-Eastwas equivalent to 70 percentof the country'stotal exportsof goods and
non-factorservices(Table 1). More recently,the slowing down in economic activity
in the major labour-receivingcountries togetherwith increased competition from
other labour-exportingcountries has led to a decline in the outflow of migrant
workers,and, with the quickeningpace of returnmigration,thereis a decline in the
stock of Pakistaniworkersin these countries.2 This has put considerablepressureon
the domestic employment situation.While remittancesfromthe Middle East coun-
trieshave also declined afterreachinga peak in 1982-83, the absolute decline has not
been substantial,although there is still considerable apprehensionabout the volume
of the remittancesin futureyears.
While there is considerable uncertaintyas to what the futuretrendsmay be,
thereis a broad consensus that the peak of the Middle East 'boom' is over and that
the migrationwhich started in 1976 will never reach the same intensityas it did
duringthe 'decade of migration',especially between 1976-77 and 1982-83. Giventhe
fact that this migrationhad such a far-reaching effecton the entire socio-economic

*The author is Senior Development Economist at ILO/ARTEP, Asian Development

Programme,New Delhi (India).
1
Hereafterreferredto only as the Middle East.
Accordingto officialestimatesof out migrationthe total numberof workers(with over
90 percentforthe Middle East) declined from168,403, in 1981 to 88,461, in 1985. There are no
officialestimatesof returnmigration.However,two one-monthsurveys,coveringall internation-
al airports,carried out in the ARTEP Phase II MigrationProject in May and October 1985,
showed the number of returnmigrantsto be, on the average, 1 1,608 per month. For a discus-
sion on the factorsleading to a decline in Pakistan'sshare in the stock of workersin the Middle
East, see [8] . The resultsof the ARTEP Phase II MigrationProjectare to be publishedshortly.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
758 RashidA mjad

Table 1
' Remittances
Workers fromMiddleEast and BalanceofPayments
(Percentages)

Workers' Workers' Workers'Remit-

Remittancesas Remittancesas tancesas % of Debt Service
Year % of Exports+ % of Imports+ TradeBalance Payments as
Non-factor Non-factor + netNon- % of Workers'
Services Services factorServices Remittances
1976-77 30.91 15.09 29.46 71.89
1977-78 56.61 28.30 56.58 35.58
1978-79 52.02 24.44 46.09 39.78
1979-80 46.13 23.87 49.47 42.85
1980-81 48.33 25.78 55.25 86.11
1981-82 60.62 27.70 51.01 26.54
1982-83 70.47 36.48 75.61 26.38
1983-84 68.14 33.25 64.95 31.02
1984-85 63.84 29.14 53.62 38.09
1985-86 - - - 39.49
Source:Table4.

structure of the country,its slowingdown,too, is boundto haveimportant reper-

cussions.The impactof thisslowingdown on the domesticemployment situation
and on the functioning of the domesticlabourmarkethas been the subjectof a
number of recent studies,and a broadconsensus, at leastas regards thelatter,seems
to be now emerging.Duringthe initialspurtin migration, the concentration of
demandforparticular skills,especiallyin theconstruction sector,led to severeshort-
ages in thedomesticeconomy,andrealwagesof workers in construction and specific
skillsin manufacturing increasedsubstantially. However, over the lastfew years,as a
result of both the increase in domestic supply(including public-sector investment in
skill-development programmes) and the slowingdown of overseasdemand,these
shortages havebeen satisfactorily met,and the resultsof two recentsurveysclearly
show thatfirmsin manufacturing and construction do not findcapacityutilization
constrained as a resultof skillshortages causedbymigration to theMiddleEast.3

3The threeearlierstudies which analysed the impactof migrationon domesticwage rates

and skillshortagesin the manufacturingand constructionsectorwere Guani et al. [3] , Irfan[10]
and Akmal Hussain [5] , the resultsof which were includedin ILO/ARTEP [7] . The resultsof a
recent field surveyon labour marketadjustmentto immigration in Pakistan,conducted in 1985,
are published in [13]. A survey of firms in the manufacturingand constructionsectors,
conducted under the ARTEP Phase II MigrationProject at the end of 1985 and early 1986, also
broadly confirmsthe resultsof the ManpowerDivision [13] . As mentionedearlier,the resultsof
ARTEP Phase II are to be publishedshortly.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 '
WorkersR emittancesand Pakistan's Economy 759

As to the impactof remittances on the domesticeconomy,thishas been

limitedeitherto its overallbalance-of-payment s effector to studiesbased on field
surveys which have analysed the use of remittancesby householdsof migrant work-
ers.Thereis littleworkon Pakistanat leastas faras theauthoris aware,on trying
to combinetheoverallmacroimplications ofthesemicrostudiesand to evaluateand
analyse the impact of the use of workers'remittanceson theoverallgrowthof the
economy as well as on its important sectors.4 one findsthatthedevelop-
Similarly,
mentplanningprocessin the countryhas rarelytakeninto accountworkers'remit-
tanceswhenanalysingoverallinvestment, domesticsavingsand sectoralinvestment
andgrowth.5
This paper makes a verypreliminary attemptto fill in this gap. Its main
purposeis to clarifyconcepts, identifyissuesand pointout theglaringdata gaps
whichexist. However,on the basis of availablemacrodata as well as theresultof
previousand someveryrecentmicrostudies,thispaperdoes tryto relatetheuse of
remittances by the migrant householdwiththeoveralldevelopment of theeconomy.
In thecourseof theanalysis,certainpolicyissuesalso emergeand thesearesuggested
as possibleareasof government intervention.

I. WORKERS' REMITTANCES AND NATIONAL

INCOME ACCOUNTS

Pakistanis amongstthe few countrieswhichincludeworkers'remittances

separately estimates[19] . A convenient
in theirgrossnational-income startingpoint
is, therefore,to examinethe basicnational-income-accounting for,as we
identities;
shallsee, thisprovidesimportant intohow workers'remittances
insights affectthe
overalleconomyand,especially, suchkeyvariablesas thesavingrate.

Net FactorIncomefrom = (InvestmentIncome+ Workers'Remittan-

Abroad ces + OtherPrivateTransfers
receivedby
fromabroad) - (Investment
residents In-
+
come PrivateTransfers made to non-
residents) (i)

4
There are, of course, a large numberof studieswhichhave triedto evaluate the effectof
migrationon the developmentprocess includinga number of country-specificcase-studies,of
which Paine*sstudy [161 on the Turkishmigrantworkersis well known. For an excellentreview
of the developmentliteratureon the subject, see Irfan [10] and Azfar Khan [11]. For a study
on the politicaleconomy of manpowerexportsfromPakistan,see JamilRashid [18].
5This
is, farexample, most glaringlytrue of Pakistan's Sixth Five Year Plan,as well as of
most officialanalysisof Pakistan'seconomic performancepublishedby the PlanningCommission
or the Ministryof Finance.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
760 RashidAmjad

Gross National Product = Gross Domestic Product + Net Factor In-

(at marketprices) (at marketprices) come fromAb-
road (ii)

Gross Domestic Product = Total Consumption + Total Investment

+ Exports (G&NFS)* - Imports(G&NFS). .(iii)

Balance
External-resource = Imports(G&NFS) - Exports(G&NFS) . . . (iv)

Total Investment** = Gross Domestic Savingsplus ExternalRe-

source Balance (v)

Gross Domestic Savings = Total Investment - External Resource

(taken as percentof GDP) Balance (as percentof GDP) (vi)

Gross National Savings = Gross Domestic Savings plus Net Factor

(taken as percentof GNP) Income fromAbroad (as percentof GNP). .(vii)

Total Investment = Gross National Savings plus External Re-

source Minus Net Factor Income from
Abroad (viii)

♦G&NFS = GoodsandNon-factor
Services
**Y=C+1+X-M
orI=(Y-C)+(M-X)
=S+(M-X)
In the above identities, the important distinction is between national and
domestic savings and the financingof investment.As Equation (v) shows, total
investmentin the economy may be financed through domestic savingsor through
externalresources.The external resourcesavailable are representedby the external-
resource balance, i.e. Equation (iv), which shows the physical resourcesmade avail-
able by foreignloans and grants,net factorincome fromabroad and other financial
transactionsincluding the use of reserves.Gross domestic savingsis thereforetotal
investmentminus the external-resouree balance, Equation (vi), and is normallycal-
culated as a residualand expressedas a percentageof GDP.
Now, duringa time period when net factorincome fromabroad is not signifi-
balance can
cant (as in Pakistan duringthe Fiftiesand Sixties), the external-resource
be taken as a fairlygood indicatorof the foreignloans and grantsavailable to finance
the excess imports over exports and this, togetherwith domestic savings,finances
total investmentin the economy. However, duringa time period when net factor

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 ' R emittancesandPakistan
Workers 'sEconomy 76 1

income from abroad (including workers' remittances) is significant,the external-

resource balance reflects,in terms of financingthe deficit,both foreignloans and
grantsand workers'remittanceswhich are available in foreignexchange. To take into
account these foreign-exchangeresources which are now made available to the
domestic economy, the identityused is that of gross national savings,to calculate
which we add gross domestic savingsto net factorincome (includingworkers'remit-
tances) fromabroad [Equation (vii)] . In termsof the identitiesdefined earlier,this
basically means that to the extent the external-resourcebalance is being financedby
net factor income from abroad (including workers' remittances),the reliance on
sources like foreignloans and grantsto finance total investmentis decreased and
thereis consequentlyan increasein grossnational savingsby thisamount.
As regardsthe level of consumptionand savings,the grossdomestic savingrate
basically reflectsthe differencebetween total consumptionin the domesticeconomy
(includingthat financedby workers'remittances)and domestic income, i.e. exclud-
ing workers' remittances.A decline in the gross domestic savingsthereforeimplies
that the reliance on outside resources(in this case, workers' remittances)to finance
domestic consumption has increased. However, by including workers' remittances
in GNP and, hence, national savings,we have a measure of total consumption and
total income withthe differencereflectingnational savings.
The importantpoint to emphasize is that workers'remittancesfromoverseas
as part of the net factorincome fromabroad basically servestwo primarypurposes.
Firstly,it supplementsthe foreign-exchange resourcesavailable to the economy and
to that extent it reduces the balance-of-paymentsconstraint. Secondly, the corre-
spondingdomestic resourceswhich are generated(i.e. the Rupee resources)can then
be used to supplement domestic investmentor domestic consumption. However,
the overalleffectof the total resourcesavailable throughworkers'remittanceswill be
finallyreflectedas an increase in total imports,equivalent to this amount. It is for
this reason that the basic identityused for total resourcesavailable to the economy
is shown as

Total Resources = GNP (at market prices) + Imports (G&NFS) - Exports

(G&NFS) - Net Factor Income from Abroad = Total
Investment(Gross Fixed Capital Formation + Changes in
Stocks)+. Total Consumption(Private + Public).6

While some of the above relationshipsmay not be intuitivelyobvious, theirsignifi-

cance will become clearerin the analysis that follows.

6 These identities are used to estimate total resources in Pakistan national income
accounts,see Economic Survey1985-86 [12) , StatisticalAnnexure,Table 2.2 and Table 2.4.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
762 RashidAnyad

IL WORKERS' REMITTANCES FROM THE MIDDLE-EAST

AND THE NATIONAL ECONOMY

We now examinein somedetailtheperiodafter1976-77whenworkers'remit-

tancesfromMiddleEastcountries beganto havea significant impacton thedomestic
economy. These countriesincludeU.A.E., Libya,Bahrain,Kuwait,Qatar,Muscat,
Oman,Iranand SaudiArabia.
Tables 1 to 7 bringout boththeimpactof remittances in termsof balance-of-
paymentssupportand the behaviourof theoveralleconomyand its majorsectors
duringthis period.We may summarizethe main featuresthatclearlyemergeas
follows.
(i) Firstand foremost is the dramaticsixfoldincreasein remittances fromthe
US
MiddleEast, from $434 million 1976-77 in to a peak US $2344 millionin
of
1982-83,afterwhichit declinedalthoughtherewas a veryslightincreasebetween
1984-85and 1985-86(Table 2).
(ii) As a percentageof GDP, workers'remittances fromthe Middle East
increasedfrom3.16 percentin 1976-77to a peakof 9.39 percentin 1982-83.After
this,theydeclined,butin 1985-86theircontribution (7.01 percentof GDP) was still
extremely significant.
It is important to note thatin our estimatewe are only including workers'
remittances in cash and not in kind(i.e. consumerdurablesbroughtalongby the
workeron his finalreturnor visitshome).Thenational-income -accounting estimates
of net factor-income fromabroad forthe periodfrom1976-77to 1982-83do in-
cluderemittances bothin cashandin kind.Separateestimates of theamountsshown
forremittances in kindare not available.However,a comparison of theestimates of
total workers'remittances and net factorincomefromabroad duringtheseyears
showsthe difference to be marginal.The generalimpression, as wellas someof the
evidencewe shallpresentlater,will showthat the amountbroughtby workersin
theformof consumer durablesin kindwas quitesubstantial. A case forincluding this
itemin net factorincomefromabroadmaybe justified.However,thedifficulty in-
volvedin obtainingor forming somereliableestimateof theseflowsmaybe a major
reasonof its discontinuation after1982-83.However,to maketheseriesconsistent,
at leasttheamountattributed to remittancesin kindshouldbe separately available.
(iii) Tables 1 and 4 bringout thecontribution of remittances fromtheMiddle
East to the balanceof paymentsincludingdebtservicerepayments. At its peak in
1982-83remittances fromthe MiddleEast contributed as muchas 75 percentto the
overallbalanceof trade(includingnon-factor services)and financed36 percentof
themerchandise imports and non -factor In 1982-83,debtservicepayments
services.
wereonly26 percentof workers remittances fromtheMiddleEastas comparedwith
almost72 percentin 1976-77.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 'Remittances
Workers 'sEconomy
andPakistan 1 63

Table 2
BroadMacroIndicators
(MillionUS dollars)

Workers'
GDP Total Workers'Remittances Net Factor PerCapita
Year (at factor Remittances fromMiddle Incomefrom Income3
cost) East1 Abroad2
1976-77 13736 578 434 554 211
1977-78 16154 1156 933 1226 249
1978-79 17984 1397 1096 1468 268
1979-80 21273 1748 1363 1847 314
1980-81 25010 2117 1667 2292 362
1981-82 27472 2225 1850 2403 381
1982-83 25584 2886 2403 3090 353
1983-84 27611 2731 2344 2933 369
1984-85 27769 2446 2069 2504 351
1985-86 29941 2665 2Q994 2683 365
Source: [12, StatisticalAnnexure]. FiguresforGDP, Net FactorIncomefromAbroadand per
Capita Incomewereconvertedto US dollarsat theofficialexchangerateas givenin the
same source(p. 141). Exchangeratefor1985-86is forJuly1985-March1986. Figures
for remittancesfromthe Middle East forearlieryearsare fromthe Government of
Pakistan'sPakistanEconomicSurvey(variousissues).
Notes: l IncludesU.A.E., Libya,Bahrain,Kuwait,Qatar,Muscat,Oman,Iranand Saudi Arabia.
2The estimatesof Net Factor Income fromAbroad for the years from 1976-77 to
1982-83includeremittances bothin cashand in kind.
3GNP (at marketprices)dividedby population.
4
Extrapolatedon thebasisof July-Marchfigures.

Table 3
BroadMacroIndicators

PerCapita Remittances
Year Remittancesfromthe fromtheMiddleEast as %
MiddleEast as % of GDP of GNP per Capita1

1976-77 3.16 2.77

1977-78 5.78 4.9
1978-79 6.09 5.19
1979-80 6.41 5.32
1980-81 6.67 5.58
1981-82 6.73 5.62
1982-83 9.39 7.65
1983-84 8.49 6.91
1984-85 7.45 6.21
1985-86 7£1 5£9
Source: Table 2.
l fromM.E. dividedby totalpopulation.LatterGNP percapitaderived
Note: Total remittances
from[12].

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
764 RashidAmjad

Table4
Workers fromMiddleEastand B.O.P.
Remittances
(MillionUS dollars)
Workers'
Remit- Debt
tances Merchan- N.F.S.1 Merchan- N.F.S.1 Service
Year from dise Exports dise Imports Repay-
Middle Exports Imports ment2
East
1976-77 434 1132 272 2418 459 312
1977-78 933 1283 365 2751 546 332
1978-79 1096 1644 463 3816 669 436
1979-80 1363 2341 614 4857 853 584
1980-81 1667 2799 650 5563 903 602
1981-82 1850 2319 733 5769 910 491
1982-83 2403 2627 783 5616 972 634
1983-84 2344 2669 771 5993 1056 727
1984-85 2069 2475 766 6009 1091 788
1985-86 2099 27023 (n.a.) 58303 (n.a.) 829
Sources: (i) [12]. Merchandiseimports and exports from Annexure Table 10.1 and Debt
servicerepaymentsfromTable 1 1.3, p. 148.
(ii) For Non-FactorServiceexportsand imports,
World Bank. Pakistan Economic and Social Development Prospects Volume 1.
(Report No. 5962-PAK) Washington,D.C. February1986. p. 133.
Notes: 1
Non-factorServices.
2
Debt service paymentsare exclusive of charges on I.M.F. facilitiesand short-term
borrowing.
3
Extrapolatedon the basis of July-Marchfigures.
n.a. = not available.

(iv) Despitethe substantial

increasein thelevelof importsmade possibleby
thereis no significant
remittances, changein thelevelof investmentas a percentage
of GNP (at marketprices)duringthisperiod.As Table 5 shows,thelevelsof total
investmentin the economyincreasedverylittlecomparedwiththosein the earlier
period from 1970-71 to 1976-77and were substantially loweras comparedwith
thosein the firsthalfof the Sixties.Thiswouldbe suggestiveof the factthatthe
proportionof incomefromabroad,including workers'remittances,thatwentinto
investment was not much different fromthat whichwent in investment from
domesticincomein thisor theearlierperiod.
(v) As regardsthelevelof savings,as Table 5 shows,whereasgrossdomestic
savingsdeclinedsharplyduringthe periodfrom1971-IS to 1985-86as compared

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 ' R emittances
Workers 'sEconomy
andPakistan 765

Table 5
Investment
and SavingRates

as
TotalInvestment GDS as Percent GNS as Percent
Years Percentof GNP of GDP ofGNP
(marketprices) (marketprices) (marketprices)
1960-61to 1964-65 18.04 10.02 10.06

1965-66to 1969-70 16.09 12.42 12.38

1970-71to 1976-77 15.05 11.04 12.04

1977-78to 1985-86 16.03 7.09 14.26

Source: Computed from[12, StatisticalAnnexureTable 2.3 (p. 21)] .

withthosein theearlierperiods,thelevelof grossnationalsavingsincreasedsignifi-

cantlyand was doublethe domesticsavingrateduringthisperiod.Whatthisclearly
shows(as we have seen in Part I) is thatthe increasein workers'remittances de-
creasedsubstantially the dependenceon foreign borrowings to financetotalinvest-
ment.Thishappenedto an extentthatdependenceon externalborrowing declined
to 12.5 percentduringthisperiodas comparedwith45.3 percentin thefirsthalfof
theSixtiesand 22.3 percentduringtheperiodfrom1970-71to 1976-77.
It is also importantin thelightof our discussionin PartI to interpret
thede-
cline in grossdomesticsavingsduringthisperiod,whichfell from11 percentin
1970-71to only7 percentin thesubsequentperiod.Whatthefallin domesticsaving
ratesbasicallyshowsis the proportion of domesticconsumption thatis now being
financedthroughworkers'remittances and not & sharpshiftupwardof thelevelof
consumption as a proportionof grossnationalincome.If,forexample,we calculate
the averagepropensity to consume(ape) between1970-71 and 1985-86,thereis
hardlyanychangeoverthe period.In fact,takingtheaveragefortheyearsbetween
1976-77and 1985-86,the ape is 0.86 comparedwith0.88 between1970-71and
1975-76.7
(vi) Giventhatthelevelof investment didnotincreaseduringthisperiod,the
highrateofgrowth(6.8 percent)of GDP duringtheperiodfrom1976-77to 1985-86
as comparedwith3.8 percentin theearlierperiodoftheSeventies(Table 6) reflects
itselfin a drasticdeclinein the incremental capitaloutputratio(Table 7). The
declinein the capital output ratio after 1976-77 is indeed so steep,from3.57
between1970-71and 1975-76to 2.5 between1976-77and 1985-86,thatit maycast
7
Estimates of gross national income and total consumption are from [12, Statistical
AnnexureTables 2.2 and 2.3 ].

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
766 RashidAmjad

.2 (N ЧО •-; О <Ч Г^
£ ^t ^t r^ v¿ ЧО vr>
&

lo 'Я «
13 P¿ ¿< 00* l> Tt Г^ 00 t^
I *
t: * -à e
^äig о » Q o> ^ ^ о
2 2 ^
S '■? О *o ro oo oo oo
22 i S
h «я c3

2 С <N О. ГО (N ГО 00
^.SoŇ vî чо oô r^ on

1
^<£f>ON OS СП Tt ^ Tt
^g(N (N t^ On Ол O'
I
33 €
3 Os СО <N ^
í ,M oq on чо ^ <n ^-^
So 2g ^ o' <n os о oô ¿

-
S3 Q} т-н (SJ 00
I
.ä g oq го чо os os oqSi

PLiVh00. ^ °) °°. °°. °) '3

QO'¿ чо rô чо ЧО чо 5
О о ^

m о r^ чо го чо §
чс r^ r^ oç oç 09 ¿j
^ On ЧО «^ ÍN m T-j
ЧО ЧО Г- 00 00 00 «
*л On On On On On On "S
£^н ^н ^н ^ ^н -g,
«о
^Н -«-•
о
4->
о
-«->
о
-*-»
о
4-»
об
^_» Q
О in О Г^ С"^ fO U
чо чо г^ гр г^ оо -
On Tf On ЧО ЧО (N ^
щ ЧО ЧО Г^ t^ 00 ^
On On On On On On §

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 ' 1 67
WorkersR emittancesand Pakistan's Economy

Table7
Capital- OutputRatio1
Incremental

1960-61to 1964-65 3.04

1964-65to 1969-70 2.64

1970-71to 1975-76 3.57

1976-77to 1985-86 2^50

Source: Computedfrom[12, StatisticalAnnexureTable 2.4 (pp. 23-24)] .
Note: l Increase in GDP (at constant factor cost) divided by Total Investment(Gross fixed
Capital Formationplus changesin stocks).

doubton theestimatesof eitherinvestment or of outputgrowthin thelastperiod.

However,it is interesting to findthatthelow capital-output ratioduringthisperiod
is about the sameas in thesecondhalfoftheSixtieswhichalso followeda periodof
a highcapital-outputratioin the firsthalfof the Sixties,althoughthiswas not as
highas fortheperiodfrom1970-71to 1976-77.
As to theveryhighrateofgrowthof domesticoutputin thisperiod,a number
of factorscertainlycontributed.Firstly,the declinein the capital-outputratio
reflectsthecomingon streamof a numberof long-gestation projects(e.g. steelmill)
whichstartedin the yearsbefore1976-77.It also reflects bettercapacity-utilization,
especiallyof themanufacturing sector,madepossibleby a moreliberalimportpolicy
and betteraccessto imported rawmaterials.1 Thiswouldcertainly nothavebeenpos-
siblewithoutthe substantial support to the balance of payments byworkers'remit-
tances,which made it possible to finance a part
significant of the tradedeficit.In
agriculture, of
a combinationof good weather,timelyavailability inputs,price
incentives and,veryrecently, in the case of cotton,a majorbreakthrough in produc-
tivityas a resultof pesticideuse, all contributed to a highrate of growthofoutput
duringtheoverallperiod,despite individualyears of bad harvests.Thiswas a major
changerelative to theearlierperiodof theSeventies.8

III. USE OF REMITTANCESBY HOUSEHOLDS AND

IMPACTON DOMESTIC ECONOMICDEVELOPMENT
Whilethe contribution to the balance of paymentand gross
of remittances
nationalsavingsclearlyemergesfromthe foregoing analysis,what has not been
is
investigated
specifically its impacton sectoralgrowthrateof outputand invest-
ment.A majorhurdlein undertaking such an exerciseis thelack of reliabledata.

8
For a comparisonof overalleconomic managementduringthe two periods,see Ahmed
andAmjad[2].

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
768 RashidAmjad

There are basically three kinds of problems. Firstly,reliable data as regardsgrowth

of output and investmentare not available forthe sectorswhich are most affectedby
remittances(e.g. small-scalemanufacturing, construction,ownershipof dwellingsand
retailtrade). Secondly,veryfew studiesexist on the patternof use of remittancesby
migranthouseholds. Also, the few studies that do exist were carried out in the very
early years of migration. Thirdly,it is generallyacknowledgedthat a significantpro-
portion of remittancescome to the country throughillegal channels, mainly the
hundi system.9 Therefore,while workers'remittancesofficiallyrecordedreflectthe
foreignexchange available to the country,these may not be a good indicatorof the
total amount of domestic (i.e. rupee) resourcesavailable to migranthouseholds.
Withinthese extreme data constraints,what follows is a rathersimpleattempt
to relate the macro with the micro studies in order to get some feel of the issues
involved,mainly to place us in a position to make suggestionsas regardsboth further
investigativework and data needs formakinga meaningfulexercise possible. For the
present analysis, for estimates of sectoral growthof output and investmentwe use
the officialstatisticsavailable and these are shown in Tables 6 and 8. Since the flow
of remittancesslowed down after1982-83, we have separatelyshown the growthrate
of output formajor sectorsforthe periods from1976-77 to 1982-83 and from1982-
83 to 1985-86. The impact of the slowingdown of remittanceswould, of course, be
with a lag. It is, therefore,too early to expect a significantchange. However, the
period of the last three years may give some indicationof this. As regardsthe use of
remittances,we have relied primarilyon the results obtained by Gilani et al. [3]
based on the field surveyof migranthouseholds carried out in 1979. These findings
have been supplementedwith some verypreliminaryresultsfromthe ARTEP Phase
II Migrationstudy based on a sample of 1360 households coveringboth rural and
urban areas in the country excluding Baluchistan province.10 As regardsthe third
problem mentioned earlier, the divergencebetween 'actual' and 'official' estimates
of remittancesis extremelydifficultto measure. Again, some preliminaryresultsof
ARTEP Phase II MigrationProject provide us with an estimate. Based on the stock
of workers abroad, their skill composition, average earningsand remittances,the
total remittancesthat would have been sent back by workersto their families,in-
cludingcash on visits,is estimated.This exercise, which was carriedout for 1980 and

9The hundi an
is informalbill of exchange by which the migrantreceivesthe equivalent
of rupee resourcesin lieu of his foreignearningsby an individualagent or firmoperatingabroad.
The foreignexchange resources,thereforedo not formallyenter the country'sbalance of pay-
ment account. However,a portionof thisamountis used to financesmuggledconsumerdurables
and goods into the country as well as to finance consumption and investmentsabroad by
Pakistaniresidents.
The surveywas carried out in end-1985 and early 1986. The major objective of the
surveywas to understandthe process of re-absorptionof the returnmigrantsinto the labour
marketand the economy. The detailed resultsof the survey,includingsamplingdesign,are to be
publishedshortly.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 Workers
'Remittances 'sEconomy
andPakistan 769

R^^hONCOONOOOOQOOCO
e rooicot^ON^coinvooN

«I
Ve'

-«-» ев
8 g Ю OÌ 00 О «О Г- СО ON Ш *Г>

I
5а*р^ CN СП г-н 00 Г^ ON СЛ -ч 00
^ - "• CN ^н ^н г-н О1СП СО
Sg^4

■=!
1ч»

I
'S. S On «Л СО СО О *-н On ^ПЧО »П
оЗдО СО Г- Q *П О ON ON^ř О
^а>г^ о (N о оо »о оо coon г^
д^^н <N CN СО СО ^ Ю VON© Г^

1
00 -S

II ^^00
д^^п
СО
vo
СО
r*
Ol
on
Ä
o
OÌ
oí
^
co
Г*-
Tř
On
»n
^
r^

1
»
I
"^wi40 ^ W^ ^ ^ Oí 00 «П On Г*-"
<£r Oioou^ONONirivo*nvooo»o
йч^»П ^- r^ *n Ol Ol O VO ^ CO<N •
C1^^ ^н ^н CM CO CO Tf U4 Г^ ON

I
Зтои^тооог-ооооооЗ
On 00 CO vO *-* 00 Г*** Th ©
■í CO О
gO^-^^-vovOTft^^OÉ

^
•CCOCOCOTÍ-^^^vot^ooS

1
1
St^OOONO^OlCO^^vO
■ N
>и Г7 t^ г^ oç 00 09 00 00 oç 00
VOr^OOONO^CNCO^t«nu
r^r^r-r^ooooooooooooa
OnOnOnOnOnOnOnOnOnONo
4

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
770 RashidAmjad

1985, showedthatofficialremittances maybe as low as 57 percentof totalworkers'

remittances sentfromabroad.11 However,at that stage,sincethe ARTEP survey
resultsof returnmigrantswas not available,theestimatesused foraverageearnings
werebasedon Gilani's1979 results,to whicha nominalincreasein wageswas added
for the 1980-85 period.The resultsof the recentARTEP surveynow tend to
suggestthatthismay have overestimated averageearningsin 1985. Whiledetailed
alternative havenot yetbeencalculated,roughestimates
projections suggestthatthe
difference between'actual' and 'official'remittancesmay be closerto 25 percent
thanto 30 percent.For thepurposeof thepresentexercise,we haveusedthefigure
of 25 percentto indicateremittances whichcome into the countrythrough'un-
official'channelsand to indicatethe 'maximum'remittedresourcesavailableto
householdswithmigrants in the MiddleEast between1976-77 and 1985-86. The
official
estimatesareusedto indicatetheminimum figure.

AverageEarnings,Expenditure,Remittancesand
Savingsof MigrantsAbroad
We startby examiningthe total earningsof a migrant workeroverseas.This
maybe dividedintothreemainparts.The firstis theconsumption expenditure incur-
redduringhis stayabroad.The secondis theamountremitted by themigrant to his
family from abroad.The which
third, is theresidual,
may be taken forthe timebeing
as his savingsabroad. Thesesavingsmaytheneitherbe used to purchaseconsumer
goods and durableswhichhe maybringalongwithhimon visitsor on finalreturn
(for his personaluse, for membersof his householdor forgiftsforfriendsand
relatives),or broughthomein theformofcashsavings, againon visitsor finalreturn.
The total cash remittances receivedin Pakistanwould be the sum of remittances
sentfromabroadand cash broughtby themigrant on hisvisitsas wellas on hisfinal
return to Pakistan.
Estimatesof averageearnings,averageexpenditure,averagesavingsand average
remittances by theworkerwhileabroad,as shownby the resultsofGilanieřя/,and
ARTEP PhaseII study,aregivenin Tables9 and 10. Gilani'sestimates werebasedon
interviews withhouseholdsof themigrant whowere,in thevastmajority
workers, of
cases,stillabroad.The ARTEP resultsare based on interviews withreturnmigrants
who havefinallyreturned to thecountry.Also,whileGilani'sresultsarefortheyear
1979, the ARTEP resultsare the averageforthe periodtill 1985 and mainlycover
theyearsbetween1975 and 1985.

11
A preliminaryreport [11], entitled 'Estimates of Demand for Pakistani Labour and
WorkersRemittancesfromSelected Middle East Countries(Saudi Arabia, UAE, Kuwait, Oman
and Qatar) 1986-90' as part of the ARTEP Phase II MigrationProject was submittedto the
PlanningCommissionin March 1986.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 Workers'
Remittances
andPakistan
'sEconomy 111

Table9
Gilani'sEstimatesofAverageEarnings, Remittances
Expenditure,
and SavingsofMigrantWorkers
(Thousand Rupees)

AnnualEstimates
for
Serviceand
Unskilled Skilled Clerical
Workers Workers Professional Workers Businessmen
AverageIncome 45.06 53.8 117.6 60.16 77.92

Average
Expenditure 13.31 19.44 31.52 20.8 28.71
(29.5) (36.2) (26.8) (34.6) (36.8)

Average
Remittances 23.74 2834 53.68 33.84 31.94
(52.7) (52.6) (45.6) (563) (41.0)

AverageSavings 8.01 6.02 32.4 5.52 17.27

(17.7) (11.2) (27.6) (9.1) (22.2)
etal [3,p. 105).
Source:Gilani
Note: Figuresinparenthesisarepercentages.

Table 10
ARTEP Estimates(Average1975-85) ofAverageEarnings,
Expenditure,
Remittancesand SavingsofMigrantWorkers

AnnualEstimates
Rupees Percentages
AverageEarnings 58,896 100

AverageExpenditure 13 ,548 23 .0

AverageRemittances 31,068 52.8

AverageSavings 14,280 242 .

Source: ARTEP Phase II MigrationProject(PreliminaryResults).

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
772 RashidAmjad

Unfortunately,we do not have the breakdownof averageearningsfordifferent

categoriesof workersfromthe ARTEP survey(although these will soon be available)
and Guani does not provide an aggregateestimate of average earningsand expendi-
ture, although he does have an estimate of average remittances(including cash on
visits) and this comes to Rs 28,966 [3, pp. 100-101] . However,the vast majorityof
workersbelong to the skilled and semi-skilledcategoryof productionworkers12and
we may use Gilani's estimates for these categories of workers with the results of
ARTEP survey. Before this is done, it is importantto emphasize that Guam's esti-
mates of average earningsabroad are based on data provided by the head of the
household of the migrantworkerwhile those of the ARTEP are based on a survey
of the returnmigrants.Secondly, Gilani's estimates of average remittanceinclude
cash broughtby the migranton his visitsto Pakistan,while in the ARTEP estimate
it is included in the savingsestimate.Finally,and perhapsmost importantly,Gilani's
estimates of remittances would underestimate 'actual' total remittancesby the
amount of savingsbroughtby the migrantworkeron his finalreturn. This may have
considerable significance,especially for a period when permanentreturn-migration
may be extremelylargeas compared withthat forearlieryears.
However, despite these differences,the resultsof the two surveysare not very
dissimilar. The sum of average remittancesand average savings out of average
earnings comes to 77 percent in the ARTEP survey and to about 70 percent in
Gilani's survey. The averageexpenditureis significantly lower (23 percentof average
earnings)in the ARTEP survey than Gilani's average over 30 percent. It is this
of
differencewhich,togetherwiththe cash remittancesbroughtby the migranton visits
to Pakistan but excluded from Gilani's estimates of savings, primarilyaccounts
forthe highersavingrate in the ARTEP surveyresult.

Use of Remittances

Turning to the crucial issue as to how remittancesare spent, Gilani's well-

known resultsare presentedin Table 1 1, accordingto which total consumptionex-
penditure was 62 percent, real-estateexpenditure was 22 percent, and investment
and financial savings accounted for 13 percent (and there was a residual of 3
percent). Of the consumptionexpenditureincurredby the migranthousehold, ex-
penditure on marriageswas 2.3 percent,on consumerdurables 2.8 percent and on
remainingitems of expenditure57 percent. Of the large percentageof expenditure
on real estate, more than halfwas spent on the constructionand purchase of residen-
tial houses. Of the investmentand financialsavingsamountingto about 13 percent
of total remittancesreceived, over three-fourthswere industrialand commercial
investments.

12Based on data of the Bureau of

Immigrationand Overseas Employment,the outmi-
grantsregisteredwith themin the skilled,semi-skilledand unskilledcategorywas on the average
over 90 percentof the total.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 ' Remittances
Workers and Pakistans Economy 113

Table11
of TotalRemittance
CrudeEstimates ExpenditurefromMiddleEast
from1976-77to 1985-86:Alternative
I (GuardEstimates)
ESTIMATEI ESTIMATEII
as
Expenditure Remit-
(Official +
(Official
Item ofTotal
Percent tances)from Unofficial
Remittances MiddleEast Remittances)
(%) (MillionRs) (MillionRs)
1. Consumption 62.19 126,294 168,391
a. Recurring
Consumption 57 115,754 154,338
b. Marriages 2.35 4,772 6,363
с Consumer Durables 2.84 5,768 7,690

2. RealEstate 21.68 44,027 58,703

a. Construction/Purchase
of
Residential
House 12.14 24,653 32,871
b. Improvement in House 2.27 4,610 6,147
с Commercial RealEstate 5.72 11,616 15,488
Land
d. Agricultural 135 3,148 4,197

3. Investment/Savings 12.95 26,298 35,065

Investment
a. Agricultural 3.3 6,702 8,935
b. Industrial/Commercial
Investment 8.21 16,672 22,230
с FinancialInvestment/Saving 1.44 2,924 3,898

4. Residual 3.18 6,458 8,611

5. Total 100 203,077 270,770

Source: [3,p. 144].

Verypreliminary resultsfromthe ARTEP surveyare shownin Table 12. As

pointedearlier,theseresultsdifferfromthoseof the Gilanistudyin thattheyshow
returnmigrants' expenditure of remittances sentback by themwhileabroad,cash
brought on visitsand cash on finalreturn. They excludethat partof the return
migrants' savingsabroad whichwerespentmainlyon consumerdurablesand gifts
for relativesand friendsas these expenditures were incurredabroad. However,
resultsof expenditure patternsincluding theseitemsis shownseparatelyin Table 12.
A comparison of Guam'sand ARTEP resultsshowthatthemajorbreakdown
of the use of remittances betweenconsumption and non-consumption expenditures
are approximately the same,withslightly over60 percentof remittances beingspent
on consumption.As comparedwithGilani's study,theARTEP surveyshowsa much
higherexpenditure on marriages- almost10 percentof totalremittances - butthis
itemalso includesexpenditure on Haj. Expenditureincurredbymigranthouseholds on
thepurchaseof realestateis lowerin theARTEP estimate, still
although significantly

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
774 RashidAmjad

Table 12

Use of Remittancesand SavingsAbroad: ARTEP Estimates

(Percentages)
Use of Remittancesplus
Item Remittances SavingsAbroad
1. Consumption 63.3 56.8
a. RecurringConsumptionplus
Durables 53.5 47.6
b. Marriages/Haj 9.8 9.2

2. Real Estates 17.3 16.4

(while abroad)

3. Investmentwhile Abroad 1.3 1.2

4. Investmenton Return 18.1 17.1

(incl. real estate)

5 . ConsumerDurables - 4.8
(purchased abroad and brought
to Pakistan)

6. GiftsfromFamily and Relatives

on Tripsand Final Return - 3 .7

100 100
Source: ARTEP Phase II MigrationProject (PreliminaryResults).

high at 17.3 percent compared with 21.7 percent in the Gilani survey. The most
important and significantdifference between ARTEP's and Gilani's estimates
concerns investment/savings out of total remittances. Accordingto Gilani's survey
results,about 13 percentof remittancesare eitherinvestedor saved by the household
of the migrant. The ARTEP survey shows that less than 2 percent of total
remittancessent by the migrantwhile abroad are investedby the household as either
agriculturalor industrial/commercial investment. To the extent that the ARTEP
estimatedoes not reflectmigrants'financialsavingskept withinPakistan,the differ-
ence with Gilani's estimate would be slightly,though not significantly,less. How-
on
ever, final return,the migrant does invest a of
significantproportion the savings
abroad, which he brings back as cash, together with those savings which were

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 ' Remittances
Workers 'sEconomy
and Pakistan 775

accumulatedforhimor by hishouseholdwhilehe wasabroad. The totalamountof

thisinvestment comesout as highas 18.1 percentof thetotalremittances.
This particularfindingof the ARTEP surveyis extremelyimportantand
providesan interesting insightintohowthemigrant workermaybe decidingon how
to spendhis earnedincome. The mainobjectives themigrant,
of whilehe is abroad,
are to supplement the incomeof the householdthroughremittances (primarily to
its
improve consumption level) and to meet the cost of what he considers are
important socialobligations, mainlymarriages (in mostcases of himselfand sisters)
and Haj (for his parents). Very littleof the savingsfromtheseremittances are
investedby thehousehold.Alsotheotherimportant goalof themigrant, whenhe is
still abroad,is to possesshis own residential house or substantially improvethe
existingpremises. This is achieved either with remittances or withcash brought
alongon visits,whichallowshimto construct, improveor,in somecases,purchasea
residentialhouse.
The main investmentdecision in the form of purchaseof agricultural
machinery landor setting
or agricultural up of an industrial or commercial unitis put
intoeffectin mostcasesafterhe returns fromabroad.Insomecases,theconstruction,
or
purchase improvement in residentialpremises mayalso be undertaken then. The
ARTEP survey of return migrants has an advantage over the Guani surveyin thatit
coversthisaspectof migrant's behaviour, whichwouldnot havebeen satisfactorily
coveredin Gilani'sestimate. It also has important policyimplications in termsof
government assistancein guiding the migrant to make an economicallydesirable
decision with his savings,broughtwith him on final return,togetherwith
accumulated savingsfrompreviously remitted income.
The ARTEP surveyalso providesestimates of theamountspentby themigrant
on purchaseof consumergoodsand durablesmainlyforhis immediatefamilyor as
giftsforrelationsand friends.Out of thetotalincomesentas remittances and that
saved by themigrant whileabroad,theexpenditure incurredon consumerdurables
comesout as 4.8 percentof the totaland giftson visitsand finalvisitsas about 1.8
percentand 1.9 percentof total income(Table 12). The majorportionspenton
consumerdurablesis on TVs, VCRs, radiosand cassetteplayers,and comesto 44.2
percentof totalamountspenton durables. Of his totalsavingsabroad,as muchas
27 percentare spenton purchaseof durablesand giftson tripshome and final
return.13

Remittances
and theDomesticEconomy
Afteranalysingtheuse of remittances
at thehouseholdlevel,we nowturnto
itsimpacton the overalleconomy.Clearly,theabsoluteamountof remittances
was

13Based on ARTEP Phase

II MigrationProjectdata.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
776 RashidAmjad

so large that they were bound to affectthe pace and structureof growthand invest-
ment in the domestic economy as well as the level and patternof imports. If we add
together the total remittanceswhich came into Pakistan from the Middle East
countriesbetween 1976-77 and 1985-86, this comes to a total of Rs 203 billion. If
we use the ARTEP estimatesto include remittancesthroughunofficialsources,this
figurecomes to Rs 271 billion.14 As a measure of the magnitudesinvolved,total
investmentundertaken by both the public and private sectors duringthis period
comes to Rs 502 billion and that for privatesector alone to Rs 192 billion.15 This
means that total remittancesthroughofficialchannelswere in absolute termsgreater
investmentduringthisperiod.
than total private-sector
A rathersimple exercise has been undertaken by us to establish some broad
orders of magnitude of the contribution of remittancesto key sectors of the
economy and overall economic development. In termsof increasein total consump-
tion expenditure,remittancesduringthe period from 1976-77 to 1985-86 (using
Gilanťs breakdown Table 11) contributedRs 123 billion accordingto EstimateI and
Rs 165 billion according to Estimate II.16 The total increase in consumption
expenditure (at market prices) during this period was Rs 356 billion.17 Of the
increase in total consumptionduringthisperiod, therefore,at least 34.6 percentwas
accounted for by remittances,and this figurecould be as high as 46.4 percentif we
use Estimate II which includes remittancescominginto the economy throughunoffi-
cial channels.18 We do not have an estimateof the amount of consumptionexpendi-
ture out of remittanceswhich went into imports. We have already noted that the
amount spent by households on purchaseof consumerdurableswas only 2.8 percent
accordingto Gilani's estimate. ARTEPs surveydoes not have a separateestimateof
this amount but does show that a significantamount of consumerdurablesis bought
by the migranton his tripshome and on his final return. The major amount of the
increase in consumption of migranthouseholds is on food items and clothes, as
shown by Gilani as well as by Abbasi and Irfan [1] . The amount spent on weddings,
however, would include consumer durables given away as dowry but could also

1
The flow of remittancesconvertedinto rupees at the officialexchange rate. As men-
tioned earlier,we have taken the flow of remittancesthroughunofficialchannelsas 25 percent
of the total flow.
15Based
on total investmentestimates given in [12, Statistical Annexure Tables 2.3
and 2.5].
Since we are calculating the increase in consumption due to remittancesfrom the
MiddleEast duringthisperiod the level of remittancesin 1976-77 is substractedfromthe total.
17Calculatedfrom[12, StatisticalAnnexureTable 2.3] .
Since consumptionexpenditureis calculated as a residualin national income accounts
and taken as the differencebetween GNPand total investment, the inclusionof unofficialremit-
tances should also be added to GNP to get the correctshare of remittancesin actual increasein
consumptionexpenditureas shown in Estimate II. This would reduce the percentageshare in
EstimateII by a smallmargin.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 ' R em
Workers ittancesand Pakistan's Economy 111

consist of those durables brought by the migrantfromabroad. At the same time,

thereis little doubt that the 'demonstrationeffect'of these consumer durables was
considerable and that the consumer durable items bought home by the migrant
stimulateddomestic demand for these items. A large part of this demand was met
by purchases abroad by Pakistanison short visitsand throughpurchase of smuggled
goods fromwithinthe country. Officialestimatesof such consumerluxuriesgreatly
underestimatetheir actual consumption.19 Some idea of the quantum increase in
the purchase of consumer durables may be formedby seeingthe numberof licences
issued for TV sets duringthese years although,as is well known, the numbersnot
registeringtheir TV sets is extremelylarge. In 1975-76, the number of TV sets for
which licences were issued was 353,992. This, in 1985-86 had risen to 1,157,804,
showing an increase of over 800,000 sets. Again, the total number of VCRs
registeredwas 146,924 in 1985-86, which would again grosslyunderestimatethe
total importsand domestic 'illegal' purchases [12, StatisticalAnnexureTable 6.4] .
The sectors whose growthof output was significantly affectedby the increased
demand generated by remittances were small-scale manufacturing,construction,
transportand communication,and wholesale and retail trade. In the case of the
small-scale sector, as Hamid [4] has shown, there was considerable increase in
demand for basic consumer goods and durables mainly by familieswho could not
affordthese goods earlierbut were now in a position to do so because of remittances.
Many new industries,therefore,came up and a large proportionof these were in the
small-scalesector. A good example was the plastic industryproducingtableware,
utensils, water-coolers, containers, and toys. The other was the engineering
industry,producing appliances such as desert coolers, washing machines, and gas
cookers and ovens in addition to traditional items like fans, sewingmachines and
bicycles, whose demand also greatlyincreased. The residentialconstructionboom
gave rise to related industriessuch as electric cables and fittings,sanitaryware and
metal fixtures. Also, the machines used in the production of a number of these
commoditieswere produced by the local small-scaleengineeringindustry. The high
(almost 10 percent) rate of growth of output of small-scale manufacturingfrom
1976-77 to 1985-86 was a reflectionof this changingtrend in demand over this
period (Table 6). However, it is not possible to say whetheras a result of a slowing
down of remittancesafter 1982-83 this sector's growthhas also slowed down. Esti-
mates of growthrate of the entire period after 1976-77 are based on the Surveyof
Small and Household ManufacturingIndustries(SHMI) of 1982-83. It is indeed un-
fortunatethat data for the small-scale sector are not regularlyavailable, with the
result that it is impossible to pick up importantchanges in its rate of growthover
differenttime periods.
The other important sector to be affected by remittanceswas construction,
which had a growth rate of over 8 percent during the period from 1976-77 to

19
A case in point are the estimatesand discussionon consumerluxuryitemsexpenditure
in[12,p.3].

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
778 RashidAmjad

1985-86. This was 30 percent higherthan the growth rate of this sector in the
precedingperiod,viz. from1969-70 to 1976-77. Accordingto the Federal Bureau of
Statistics,this sector covers all constructionactivitiesincludingrepairs,maintenance
and demolitionof buildingsand otherconstructionworksundertakenby households,
privatebodies and public authorities. Afterthe slowingdown in the inflowof remit-
tances after 1982-83, the growth of this sector could have been expected to slow
down but, on the contrary,the growthrate as givenby officialfiguresincreased to
9.8 percentin the last threeyears as compared with 7.3 percentin the six years after
1976-77. Here, again, because of insufficientstatisticalinformation,the contribu-
tion of the constructionsector is indirectlyassessed on the basis of certainassump-
tions (e.g. for urban areas the cost of constructionis ten times the availabilityof
cement forlocal consumptionand 40 percentof the total cost equals the value added
in construction. The value of the rural works programmeis taken as 60 percent
of total expenditure and for the rest of the rural areas it is assumed that the gross
rent derived from the Household Income and Expenditure Survey consists of 8
percentof the cost of constructionand the value added is 50 percent thereof) [15,
p. ix] . Again, as pointed out for the small-scalesector, there needs to be a more
carefulassessmentof the growthof output of this sectorto establishchangingtrends
in output over time.
The two other sectors which showed a high rate of growth of output after
1976-77 are 'transport,storage and communication' and the ^wholesale and retail
trade' sectors (Table 6). There is, however,in the case of these two sectors a very
marginalslowing down in the growth of output after 1982-83, the peak year for
remittances,but, again, giventhe ratherindirectway in which the growthof output
is calculated forthese sectors,it is difficultto put too much weighton this.
Unfortunately,even more difficultthan relatingsectoraloutput growthto the
use of remittancesis the task of relatingtotal privateinvestmentto the investment
undertakenby the migrantworkeror his household. Investmentstatisticsprovided
by the Federal Bureau of Statisticsare again estimatedindirectly,and the assump-
tions involvedin many cases may well lead to a wide divergencewith the actual situa-
tion. In the case of investmentby the migranthousehold, thereare also many other
difficulties,especially in relatinginvestmentin construction/purchase /improvement
of residentialhouses. Where an existing residentialunit has been purchased, this
would not be an additional asset as far as the economy is concerned. Similaris the
case of the purchase of lands which are not included in investmentestimatesof the
Federal Bureau of Statisticsforownershipof dwellings.
However, some broad comparisonsare stillinstructive.Total privategrossfixed
capital formationin ownershipof dwellingscomes to Rs 44316 million,or Rs 44.3
billion (Table 8). Using Gilani's breakdown for the use of remittancesin construc-
tion/purchaseof residentialhouse and improvementin housing,the figurebased on

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 'sEconomy
andPakistan
Remittances
Workers 779

officialremittances comesto Rs 29.3 billion,and,usingtheestimateforofficialand

unofficialremittances, the figurecomesto Rs 39.1 billion. On theassumption that
purchaseof land in the case of construction came to about 25 percentof thetotal
expenditureand addingto it a figureof 25 percentforpurchaseof the existing
residentialhouses,the estimateforinvestment financed out of remittances comesto
33 percentof thetotaland,on thebasisof thehigherestimateof remittance flows,it
comesto 44 percent.Evenon the basisof the lowerestimateit is clearthatremit-
tancesamountedto a verylargepartof investment in ownership of dwellings during
this period. Indeed,if it had not been the manytentativeassumptions thatthe
estimatesare based on, one could argue that the Federal Bureau of Statistics
estimatesof investment in ownership of dwellingmaywellbe underestimating actual
investment.
In termsof investment throughremittances in othersectorsof theeconomy
thecontribution is againindirectly estimated.In termsofinvestment in agriculture
based on Gilani's estimates,the two alternativeestimatesgive figuresof 13.1
percentand 17.4 percentof totalprivategrossfixedcapitalformation in agriculture
being financedthroughremittances.20 As regardsindustrialand commercial
investment, we mayassumethatalmostall the investment in industrywentintothe
small-scalesector. As a percentage of privategrossfixedcapitalformation in small-
scalemanufacturing thecontributions
and services, of remittances areas highas 53.5
percentand 71.3 percentrespectively, whichwould seemon theveryhighside.21
Again,one may suggestthatthe FederalBureau'sestimatesof privateinvestment
in thesesectorsmaybe considerably underestimated.
It wouldalso havebeen extremely interestingto estimatethe contribution of
remittances to investment in the transportand communication sector,especiallyas
import of vehiclesby Pakistanisresidingabroad has to be done throughfirmsin
Pakistan.Unfortunately, thesefiguresarenotreadilyavailable.

CONCLUSIONS

Despitethe ratherpreliminary natureof theexerciseundertaken in thispaper,

thereare,we feel,someimportant conclusionsthatcouldbe emphasizedevenat this
stage. Theseareas follows.
in nationalincomeaccountdoes pose
(i) The inclusionof workers'remittances
some problemsin termsof interpretation of the conventionalnational-income-
accountingidentities betweentheaccounting
used. One has to differentiate conven-
tionsand behaviouraleffects.22 The declinein domestic for
savingrate, example,

20Estimatesof total investmentare fromTable 8 and of remittancesfromTable 2.

21See
footnote20.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
780 RashidAtujad

as a resultof increasein workers'remittancesis more the resultof an accountingcon-

vention than a behavioural relationship. At the same time, the estimate of gross
national savings,which increases as a resultof workers'remittances,primarilymeans
that dependence on other external resourcesto financeinvestmenthas correspond-
inglydecreased and not that the economy saves more now than it did previously.
(ii) The contributionof workers'remittancesto the national economy comes
out veryclearly fromour analyses. Also, the resultsof the ARTEP survey,even in its
preliminaryform,comes out with some interestingresults.Firstly,it shows that a
significantamount of earningsabroad are retainedby the migrantand broughtback
with him on his finalreturnto Pakistan.This resultmay have importantimplications
as faras the flow of officialremittancesis concerned.
The fact that they, in recent years,have not declined as much as one would
expect on the basis of estimates for out- and return-migration, may well reflect
these savings which are now being brought back by the returnmigrants.As these
returnflows become larger,they will 'cushion' to some extent the expected decline
in remittances.Secondly, the factthat these savingsare broughtback by migrantson
final returnto the countrymeans that it is still not too late in the day to tryto tap
these resourcesand, withgovernmentassistance,help to channelize them forproduc-
tive use. It appears that the time when he needs this assistance most is when he
finallyreturnsand is going to take a decision to use his savingsfor eitherindustrial
or commercialor, in some cases, agriculturalinvestment.
(iii) Presentstatisticson the growthratesof the sectorsmost directlyaffected
by migrationseem to be of ratherpoor quality and would make it especially difficult
to pick up any changesin trends.The situationseems worse in the case of investment
estimateswhere it would appear that officialestimatesmay well be underestimating
the actual investmentby the privatesector fora numberof sectors.A more detailed
and thoroughexercise than the one we have been able to undertakewould, however,
be needed to establishthismore accurately.
Finally, on a positivenote, the resultsof the use of remittancestend to suggest
that they may not have been as wastefullyutilized as the generalimpressionwould
indicate. Certainexpenditures,especially weddings etc., are wastefuland could have
been less extravagant. On the otherhand, the use of earningsabroad also shows that
between 30-35 percentof the remittanceswere in factinvestedin some formor the
other. That more than half of this went into real estate was to have been expected,
giventhathousingmust be consideredone of the essentialneeds. Also, if we combine
Gilani's and ARTEP results,total investmentin agriculture,industryor commerce
may be substantial,especially in terms of official estimates of privateinvestment.
What would be more beneficialto the economy in the long run would, however,be
the investmentof this money in small-scale industrial enterprisesratherthan in
commercialundertakings.

22This
point is also ratherforcefullymade in Papanek [17] .

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 ' R emittance
Workers s andPakistan
'sEconomy 781

REFERENCES

1. Abbasi, Nasreen, and Mohammad Irfan. "Socio-economic Effects of Inter-

national Migrationon the Families leftbehind". Islamabad: Pakistan Institute
of Development Economics. 1983. (Studies in Population, Labour Force and
MigrationProject, Report No. 7)
2. Ahmed, Viqar, and Rashid Amjad. Management of the Pakistan Economy,
1947-82. Karachi: Oxford UniversityPress. 1984.
3. Guani, Ijaz, et al. "Labour MigrationfromPakistan to the Middle East and its
Impact on the Domestic Economy, Part I". Islamabad: Pakistan Instituteof
Development Economics. June 1981. (Mimeographed; Research Reports
Series,No. 126)
4. Hamid, Naved. "Employment and Industrialisationin the Small Scale Manu-
facturingSector". Bangkok: ILO/ARTEP. 1983. (Mimeographed;its results
are included in ILO/ARTEP [9] )
5. Hussain, Akmal. "Impact of Out and Return Migrationon the Functioningof
the Domestic Labour Market and its Implicationsfor EmploymentGenera-
tion". Bangkok: ILO/ARTEP. (The main findingsof this mimeographed
paper are included in ILO/ARTEP [9] )
6. ILO/ARTEP. Employmentand StructuralChange in Pakistan - Issues for the
Eighties. Bangkok.January1983.
7. ILO/ARTEP. Impact of Return Migration on Domestic Employment in
Pakistan: A PreliminaryAnalysis. Bangkok. April 1984.
8. ILO/ARTEP. Mid-TermReview of the Employmentand Labour MarketSitua-
tionduringthe Sixth Five Year Plan (Î983-88). New Delhi. March 1986.
9. ILO/ARTEP. PreliminaryEstimates of Demand for Pakistani Labour and
WorkersRemittances from Selected Middle East Countries (Saudi Arabia,
UAE, Kuwait, Oman and Qatar). InterimReport. New Delhi. March 1986.
10. Irfan, Mohammad. "Consequences of Out Migration on Domestic Labour
Market". Islamabad: Pakistan Institute of Development Economics. 1983.
(Mimeographed)
11. Khan, Azfar. "The Impact of InternationalLabour Migrationto the Middle
East on the Rural Barani Areas of the Punjab and the NWFP: General Fold-
ings of Village Surveys 1984". Paper prepared for ILO/ARTEP, Bangkok.
November 1985. (Mimeographeddraft)
12. Pakistan. Finance Division. Economic Advisers' Wing. Economic Survey,
1985-86. Islamabad. May 1986.
13. Pakistan. Ministryof Labour, Manpower and Overseas Pakistanis.Manpower
and Overseas Pakistanis' Division. Labour MarketAdjustmentto Immigration
in Pakistan. Islamabad. January1986.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
782 RashidAm/ad

14. Pakistan.PlanningCommission.The SixthFive YearPlan: 1983-88. Islam-

abad. October1983.
15. Pakistan.StatisticsDivision.FederalBureauof Statistics.
NationalAccountsof
Pakistan(Productand Expenditure) 1981-82to 1984-85. Karachi.1985.
16. Paine,Suzanne.ExportingWorkers: The TurkishCase..Cambridge: Cambridge
UniversityPress. 1974. (Occasional Paper No. 41 of the Departmentof
AppliedEconomics,University of Cambridge)
17. Papanek,G. F. "The Effectof Aidand otherResourceTransfer on Savingsand
Growthin Less DevelopedCountries".EconomicJournal.Vol. 82, No. 327.
September1972.
18. Rashid,Jamil."The PoliticalEconomyof ManpowerExport".In H. Gardezi
and J. Rashid,Pakistan:The UnstableState. Lahore:VanguardBooks Ltd.
1983.
19. United Nations.NationalAccount Statistics:Main Aggregates and Detailed
Tables,1982. New York. 1985.

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 Commentson
"Impact of Workers'RemittancesfromtheMiddleEast
on Pakistan'sEconomy: Some SelectedIssues"
Thank you verymuch, Mr Chairman,for havinginvitedme to be a discussant
this afternoon. I found this to be a very interestingpaper. It offeredsome new
insightsand new data although I would have liked to have seen more of the results
of the work under ARTEP Phase II. But, as has been explained, theyare perhapsnot
quite ready yet. I particularlyenjoyed and supportthe author's focus on the concept
of 'accumulated savings'. It helps to explain, in a largemeasure,why workers'remit-
tances rose in 1985-86 despite the accelerated flowof returnees.And it has interest-
ing implications for the economy. It suggeststhat the much-apprehendedabrupt
fall-offin remittanceflows followingthe cut-back in economic-activitylevels in the
host countriesmay well not occur because accumulated savingswill provide, to use
the author's own phrase, a 'cushion' to the economy and the externalaccount. More
importantly,it will give the economy and policy-makerstime to make adjustments
to what is a fundamentaland possibly irreversiblechange in the structureof the
balance of payments.
One of the measures that the author suggestsis that the Governmentshould
consider institutingpolicies aimed at ensuringthat these flows are attracted into
productive channels. While that is a statementfull of good intentions,the author
does not say what kind of measures he has in mind: whether,for example, he is
thinkingin terms of some kind of tax or compulsory purchase of Government
financialinstruments.Nor does he discuss the difficultiesthat the enforcementof
such measureswould entail. I would be interestedin knowingwhat ideas the author
had in mind.
On the impact analysis, which is the central part of the paper, I was a shade
disappointed. I thoughtthat the methodologythat was used, confiningthe author to
the simplearithmeticof percentagesand proportions,was a littlesimplistic.Thereare
available reasonablygood time-seriesdata on sectoralgrowthratesof output, invest-
ment and workers' remittancesand on the share of remittancesgoing to different
categories of expenditure. Given this information,the impact analysis could have
been done with more rigourand strongerempiricalunderpinnings.I recall that the
firstversionof the PIDE macro-economic model had remittancesas an argumentin
the consumption function. (It was dropped discreetly in the revised version for
reasons which remainunexplained.) Mr Amjad could have tried to re-estimatesuch

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
784 MeekalAzizAhmed

a functionalrelationship,or added remittancesas an argumentin an import and in-

vestmentfuncton.He could have also triedto correlatemore directlysectoral growth
rates and remittanceflows etc., because his simple analysis does not tell us very
much. In fact, it raises more questions than it answers. In some sectors,the impact
is actually perverse;in others the effectis rathermarginal.In the small-scalemanu-
facturingsector, Mr Amjad claims that the acceleration in growthrate appears to be
correlatedwith remittanceflows. However, I am told that thereare stillsome prob-
lems with these data; in particular,the deflatoris quite suspect and the resultsneed
to be taken with caution.
Finally, despite the paper's somewhat mixed results,the author's conclusions
are ratherbold. He claims that the notion that much of these flows have been con-
sumed ratherthan investedis exaggeratedand that remittanceshave broughtsome, if
not much, benefit to the economy. I am afraid,here I must disagree. Despite this
enormous windfall,its macro-economic impact has been minimal. Our investment-
to-GDP ratiohas stagnated;our savingrate remainsdismallylow; and our persistently
large fiscal and external imbalances are a threat to macro-economic stability. Mr
Amjad talks about Pakistan's low capital-output ratio but I assurehim that our low
capital-output ratio is not so much a reflectionof a high degree of efficiencyin
capital use as a symptomof our neglect of replacementand new investmentand the
curtailmentof essential allocations for the maintenance and improvementof our
economic and social base. Pakistan is today faced with a rapidlydeterioratingphysi-
cal and social infrastructure, as well as with capital destruction.Our past neglectof
investmentin power has resulted in serious energy-shortages and extensive load-
shedding. In the irrigationsector, there has been an alarming deterioration in the
command areas of our canals and watercourses.More than half of the water is lost
as a resultof seepage and percolation,affectingcrop productionand exacerbatingan
already serious drainage problem. Our transport system has been subject to pre-
mature road failureswith more than 88 percent of our highwaysin need of rehabili-
tation. Our social sectors continue to fare poorly. This neglectis clearlyreflectedin
our social indicators - low literacyand enrolmentrates,high infant-mortality and
low life-expectancy,poor nutritionand a high and apparentlyrisingrate of popula-
tion growth. Our industryremains highlyprotected and inefficient,and turnsout
low high-cost,non-competitiveand poor-qualityproducts,mainly for the domestic
market.
My point is simplythis. Despite the windfallaffordedby the inflowof remit-
tances, Pakistan has not improvedits investmentperformanceand has not addressed
the urgent problem of substantiallyupgrading its capital stock, strengthening its
social base and bringing down its large fiscal and -account
current deficitsthrough
improvedsavingsand export performance.Whilein the near termwe could continue
to live off our capital base, as we have done in the past, such a policy will quickly

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions
 Comments 785

producediminishingreturns and growthwillfalter.To me at least,it is clearthatby

farthe greaterpartof the inflowsof remittanceshas beenused forwastefulcon-
sumptionor investedin unproductive -
uses in realestate,goldandlow-productivity
services
sector.

& Development
Planning Division, Meekal Aziz Ahmed
Govt.ofPakistan,
Islamabad

This content downloaded from 147.126.1.145 on Fri, 27 Sep 2013 01:47:44 AM

All use subject to JSTOR Terms and Conditions