7/29/2019 107507322_20041117185522

1/42

W4S i6 oPOLICY RESEARCH WORKING PAPER 1410

MacroeconomicEffects Theexperiencef 18oiexporting countries fromof Terms-of-Trade Shocks -- 973 o 1989)suggests hatpermanent erms-of4radeshodcshavea significantThe Case of Oil-Exporting Countries positiveeffe'z on:consumptionr nvestment

N.kola - t'fora and output, particularlyofNikola Spatafora .-- ;-AdeWaeontradables. here salmostAndrew Warner -{ no effecton savings, nd -at.advEneresect on the trade.balance.There-s no evidence-6of'Dutch isease ffects.

The World BankIntrntioalEconomicsDepatrtment

IntermaionalEcoDnomicnalysiand ProspectsDivisonJanuary1995

7/29/2019 107507322_20041117185522

2/42

Poi.icy RESIEARC(I-IWORKIN;PAI'hiR 1410

Summary findingsSpataforaand Warner investigate he impact on The shocksal;o have a significantpositive effect oneconomic growthiand developmentof long-run consumption.Government consumption responds almostmovements n the external tcrms of tradc, with special twicc as strongly as private consumption.reference o the experience of 18 oil-exporting countries The shocks have no effect on savingsand adverselybetween 1973 and 1989. affect the trade and current account balances.They argue that this sample approximatesa controlled There is a significantpositive effect on the output ofcxperiment for examiningthe impact of unanticipated- all main categoriesof nontradables. But Dutchdiseasebut permanent shocks to the terms of trade. They effects arc strikinglyabsent. Agricultureandanalyze he sample econometricallyusing panel data manufacturingdo not contract in reaction to an oil pricetechniques. increase. Dutch diseaseeffects may be absent in partThey find that permanent ternis-of-tradeshocks have a becauseof policy-inducedoutput restraints in the oilstronglysignificantpositive effecton investment,which sector, or because of the "enclave"nature of the oiltheyjustify heoretically on the grounds that countries in sector, which does not participate in domestic factorthe sample import much of their capital equipment. markets.

Thispaper-a product of theInternationalEconomicAnalysis nd ProspectsDivision,nternationalEconomicsDepartment-is one in a seriesof backgroundpapersprepared n support of the analysesand scenariosn GlobalEconomic Prospects1994.Copiesof thispaperare available ree from he World Bank,1818H StreetNW, Washington,DC20433. Please ontactJackieQueen, oom S8-216,extension33740 (37 pages). anuary 1995.

ThePolicyResearch orkingPaperSeries isseminateshe idings of work miprogress o encouragehe exchangef ideasahoutdevelopmentsue An objecaeof heseriess o get he indings ut quickly, ven f thepresentationsre ess han ly polished hepapers arry he names f t authors ndshould eused ndcitedaccor&ngly. he indings.nteapretations,ndcondusionsre heauthors'oumandshould ot beattributed o theWorldBank. ts Exwtive Board f Directors. r an of its moneber ountries.

Producedby the Policy ResearchDisseminationCenter

7/29/2019 107507322_20041117185522

3/42

Macroeconomic Effects ofTerms-of-Trade Shocks

The Case of Oil-ExportingCountries

Nikola Spataforaand

Andrew Warner

7/29/2019 107507322_20041117185522

4/42

MACROECONOMIC EFFECTS OFTERMS-OF-TRADE SHOCKS: THE CASE OFOIL-EXPORTING COUNTRIES.

Nikola SPATAFORA& Andrew WARNERI. IntroductionMost developmenteconomists agree that variations in world prices are an imnportant ource of risk andinstability or developing economies. Developing counaies derive about half their export earnings fromprimarycommodities,whose world prices are extremelyvolatile. Nor is such volatilitypurely short-term:mauchdata suggest that commodity prices undergo long periods of rise and decline. To provide oneexampleof the magnitudes nvolved, he World Bank's index of non-oil commodity prices has exhibiteda trenddecline of about 1.5%per annum since 1948,cumulating o a 50% decline over 45 years. Over thenext decade, the World Bank is forecasting hat this trend will reverse and there will be a rise of about0.7% per annum. What is the impact of these long-run movements in the external terms of trade oneconomic growth and development?We think that the world has provided a natural experiment over the past 25 years that can be used toanalyze his issue. Oil-exportingdevelopingcountriesexperienceda major rise in the world price of theirmain export between 1973and 1981, followedby a smaller,but still substantial,decline between 1981and1989. Nominaloil prices rose from $2.70 in 1973 to $34.31 n 1981,and then fell back to $16.31 n 1989;deflating his by US producerprices, real oil prices rose from an index of 1.0 in 1973 to 5.83 in 1981 and2.42 in 1989. From a researchstandpoint,we think that this episode represents a fortunate opportunity tobetter understand he impact of terms-of-tradeshocks, for the following reasons.First, it seems mportant o distinguishbetween permanentand tansitory, or anticipatedand unanticipatedchanges in the terms of trade, and we believe that oil prices come closer than any other data to measuringan unanticipated permanent change. The 1973 rise was not anticipated several years before, and it wasquickly and widely perceived as a permanent feature of the economic landscape. The long decline

7/29/2019 107507322_20041117185522

5/42

page 2beginningn 1981was also not widely nticipated;t probably ook onger or it to be acceptedas a long-run phenomenon,ut at leastby 1986,whenoil prices fell by 25%, he 1981 evels were long considereda thing of the past. Hence, t is crediblehat the major long run changes n oil prices werepreviouslyunanticipatedvents hat werewidelyviewedo be permanent nce theyoccurred.Second, he timespanof the riseanddecline n oil prices s fairly ong-8 or9 years-so that t is credibleto believe hat we can observe ong-run ffects. Third, he movementsn oil priceswerelarge. Hence,there is much statisticalvariation, nd if effects are present,we shouldbe able to estimate hem moreefficiently. Fourth, erms-of-trade hangeswere so dominant or thesecountriesover this period hatdecidingwhat variables o control or,so as to avoid omittedvariablebias, shouldconstitute ess of aproblemFifth, we have both a rise and a decline n oil prices, so that we can try to detect asymmetries n theresponses. Sixth,measurement roblemsassociatedwith index numbersare less of an issuefor oilexporters, inceoil is a relatively omogeneousroduct Seventh, imultaneityetween he termsof tradeand domestic conomic ariables,whilenot alwaysabsent, s at leasta tractableproblembecause here sa fairconsensus bout he causesofthemajoroil pricemovements. urther,or manyofthe countries ndmostof the variableswe examine, t seems easonable o view he termsof tradeas predetermined.Our broad conclusion s that an examination f the data for oil exporters eads to a differentpictureofterms-of-trade ffects than can be obtained rom an unbiased ampleof terms-of-trade rticles n theliterature.We find that the impacton investment s crucial o understandinghe responseof the currentaccount o, and helong-run rowth ffectsof, erms-of-tradehocks; et he iteratures filledwithcurrentaccountmodels hatholdinvestmentonstant, nd long-runrademodels hat treat the capitalstockas anendowmentwhich s unaffected y terms-of-tradehocks. Conversely, utchdisease ffectsareexaminedextensivelyn the literature, ut wefail to findanyevidence hat heDutchdisease s a majorphenomenon.Wealso ind that he simple nsights mm he tradables-nontradablesodelare wellsupported y thedata.Finally,we find hatthe response f expenditureo terms-of-tradehocks s not very sensitive owhetherthe expenditure omes from he publicor privatesector.

7/29/2019 107507322_20041117185522

6/42

page 3The rest of the paper s dividedas follows. SectionI summarizes he relevant heoretical iterature ndempirical studies. Sectionm containsa description f the data. SectionIV sets out the econometricframework ndanalysesome mportnt econometricssues. SectionV presents hemainresults. SectionVI discusses omecriticisms.SectionVII concludes.II. TheoreticalJiterature and empiricalstudiesFor expository urposes,wefind it useful o divide he heoreticaliterature n the mpactof terms-of-tradeshocks into two broad groups. One class of models s microeconomnicn nature and stresses the differingeffects of terms-of-trade hockson differentsectorsof the economy. Examples nclude he tradables-nontradablesmodelsoriginated y Meade,Salter 1959),and Swan; he DutchDiseasemodelsfound n,say, Wijubergen 1984, 1984b)and Neary& Wijnbergen 1984),and summarizedn Corden& Neary(1982)and Corden 1984); nd the computable eneral quilibriummodelof Bruno& Sachs (1982).A second lass s concermed ith he behavior f broadmacroeconomicggregates, articularlyaving ndthe currentaccount, ndrecently as tended o stress ntertemporalssues. A partial eferenceist for thisliterature ncludesLaursen& Metzler (1950),Harberger 1950),Obstfeld 1982),Persson& Svensson(1985),andBean (1986).

A. SectoralEffectsManyof the insights rom he earlier iterature tressing ectoraldisaggregationanbe obtained, ollowingCorden,by thinldng n termsof a thee-sector,perfectlycompetitive eoclassicalmodel,with fixed otalfactorendowments.Let he firstsector 0) produce il for export; et the second, DutchDisease'sector(D)produce ll other radables, oth exportablesnd importables;nd let the third ector N) produce on-tradables. The first twosectorsproduce radables t givenworldprices,and the difference etween hequantity supplied and demanded is made up by exports or imports; the price of nontradables sendogenouslyetwrminedy the conditionhat heirsupplyequal heir demand. Assume irst that outputin eachsector s produced ya factorspecific o tluht ector,andby labor,which s mobile etween ll threesectorsand movesbetween ectors o equalize ts wage in all threeemployments.

7/29/2019 107507322_20041117185522

7/42

page 4Now et therebe an oil priceshock. The direct effect s to increase he aggregate ncomesof the factorsinitiallyemployed n 0, which in turn has two effects. First, if some part of the extra income s spent,whetherdirectlyby factorowneis or indirectly hroughbeingcollected n taxesand then spent by thegovenunent, ndprovided he income lasticityof demand or N is positive,demand or N rises. Givenless than perfectlyelastic supply. he priceof N relative o tradablesmustrise. That is, there s a realappreciationwhichdrawsresourcesout of 0 and D into N. This s thespending ffect.In addition, heprice increase raises he marginalvalueproductof factors n O so that, at a constantwage n tenns oftradables, he demandfor labor n 0 rises, inducinga movement f labor out of D andN. This is theresourcemovenen ffect.The outputof D must finallybe lower hanbefore he shock,while he outputof N couldbe higheror lower, he spendingeffect ends o increase t, the resourcemovement ffect odecrease t.Assume ow,as is common, hat he oil sectordoes not n factemployany factors hat canbe used by therest of the economy. Here, the oil sector is an 'enclave' whichdoes not participate n domestic actormarkets.There s thenonlya spending ffect,and the key mechanism f resource eallocations the realappreciation. rovided pending n nontradablesnitially oesup, outputof N must inally e higher hanin thepre-shock ituation.Assume nstead hat more han one factor is mobilebetweenat least two of our threemainsectors. Forinstance, ay thatcapital s also mobilebetweenD andN, so that these wo sectors, mployingaborandcapital n different roportions,makeup a mnini-Heckscher-Ohlinconomy.Now he resourcemovementeffect an haveparadoxicalesult. Because f thepriceshock, abormovesout of thismini-economynto0. If D is thecapital-intensivendustry,ts output ouldon balanceexpand. If N is capital-intensive,heshockcouldcausea realdepreciation.At this stage t is useful o compare he predictions f this neoclassicalmodelwith a Keynesianmodel.In the neoclassicalmodel,with factors ixed in the shortrun, and full employment etermining utputthrough he productionunctions,here s no scope or 'aggregate utput' to changemuch: he shockcanat most change the composition f output. However, n a Keynesian conomywith stickyprices, thedemand timulus an raiseaggregateoutputeven n the shortrun.

7/29/2019 107507322_20041117185522

8/42

page5Regardingonger-termupply-sideffects,he neoclassicalramework redicts hat hese hingecruciallyon the relativeprice signals. Reproducibleactors, uch as physical ndhumancapital,will accumulatein sectorswhose elative rices ise ollowinghe terms-of-trademprovementmoreprecisely,he relevantprice s a productpriceadjusted or movementsn inputprices, hat s, a valueaddedprice). As we haveseen,the relativeprices n 0 andN typically ise in the shortrun, and thus factorsaccumulate n thesesectors,expandingoutput in the longer run. By contrast, capitalwill shift out of, or be allowed odepreciatewithout eplacement,n D.

B. Macroeconomic ffectsTuning now o the macroeconomiciterature, arlyanalyses, onductedwithin heframework f thestaticKeynesianaving nd investmentunctions,ncludeLaursen& Metzler 1950) ndHarberger 1950).Thedrivingmechanisms that erms-of-tradehocks ffect eal income, ndhencesavings.Given nvestment,this determineshe evolution fthecurrentaccount.Later esearch,ypifiedby Obstfeld 1982),Persson& Svensson 1985),and Sen & Turnovsky1989), sbased on neoclassicalmodels, ypically nvolvingdynamic ptimization.Terms-of-tradehocksalter permanent ncome, ntratemporal nd intertemporalrelative rices, nd henceaffect onsumption,aving, nd investment.Oneconclusiono emerge rom hisliterature s that the effectsdependcriticallyupon whether he shock is permanentor temporary, ndanticipated r urnanticipated.Let us first consider he possible mpactof a permanent, nanticipatedncreasen the termsof tradeonconsumption. he shock ncreases oth currentand permanent ncome. All models hereforepredictanincreasen aggregate onsumption. n sectoral erms,consumption f oil s probably nsignificantor oursample, and consumptionof other tradablesmust rise throughincome effects. Consumptionofnontradablesmustequal production,whichwas discussed bove.Investments treatedas exogenous ymuchof the literature.OlderKeynesianmodelsoften postulatesimpleacceleratormechanismwhereby he shock raises expected demand and hence investment.Optimizingmodels usually conclude hat aggregateprivate investment s determined y the ratio ofTobin'sq (thepresentdiscounted alueoffuturemarginal roducts f capital) o the aggregatenvestmentprice deflator. Assuming hat capitalgoods have a strong mport content, he shock boosts domestic

7/29/2019 107507322_20041117185522

9/42

page 6investment-Schmidt-HebbelServen 1993). Yet mostmodels reat nvestment oodsas dcmesticallyproduced, nd hence obtain ar more ambiguous esult-Persson& Svensson 1985).Sen & Turnovsky(1989). For developing conomies,he formerassumption ppears o be correct.Three actorsmayhowever educe he incentiveo invest. First,profitsmaybe capturedby unions rent-sharing). Yetmanyof these countries ack an organized abor movement.Second, n OPEC countriesproduction uotasmaybe seenas a permanent eilingon oil production,eaving ittle reason o invest ntheoil sector'sproductive apacity.On theotherhand,an increase n suchcapacitymight trengthen ne'shand in the periodic quotanegotiations; lso, there is stil an incentive or cost-reducingnvestment.Finally, ncreasedwealthmayencourage onsumption f leisure,and a contractionn labor supplywillreduce the marginalproductof capital;see Bean (1986),Sen & Turnovsky 1989). Yetmanyof thesecountries ace an extremely lasticsupplyof potential mmigrantworkers.Regarding he currentaccount,Laursen& Metzlerpostulated hat the shock by raising real income,increases savings. For a given investment, his increasesthe current account. In an optimizingintertemporalramework,with domestically roduced nvestment oods,Obstfeld 1982)and Svensson& Razin 1983) howed hatwith a constant ateof timepreferenceheeconomy hould ump immediatelyto its new steady state, with no effect on the current account. If instead the rate of time preferencedecreaseswithutility, here s a transitoryise n savingsand currentaccount urplus. Onthe other hand,if investmentrequires mportsof capital machinery, t may be encouraged y the shock, leading o atemporary urrentaccountdeficit.

C. Empirical StudiesWarner 1992)examinedwhether he international ebt crisis whichbegan in 1982could explain heinvestmentdecline n 14 heavily ndebted ountries. He found that equationswhichomittedall debt-related nformation, ut ncorporatedheeffectsof fallingexportpricesand high world real nterest ates,could forecast he fall in investment.Thissuggests he importance f the termsof trade n determininginvestment.Warner(1994)proposedand estimateda microeconomicnvestmentmodel to determine he relativeimportanceor Mexico's nvestment ecline n the early1980'sof threeexplanations:heoil pricedecline,

7/29/2019 107507322_20041117185522

10/42

page 7the terriination f capital nflows, nddebt-overhang/uncertaintyffects. Usingquarterlynvestment atafor68 private-sectorndustries etween1981and 1985, e found hat hemainmirroeconomicmechanismdriving he investment eclinewas he rise n therelative riceof imported nvestment oods,and furtherthat the termsof tradedecline(drivenby fallingworldoil prices)explainsmuchof the increase n thisrelativeprice.Morley 1992) xamined tabilization rogramsn a broadsampleof LDC's. Usingpaneldata, he foundthat he termsof tradehad a significant ositive mpacton investnent and output. The samewas true ofreal appreciations,whichas we have arguedare a likelyconsequence f terms-of-trade hocks.Etherington& Yainshet 1988),using ime-seriesmethodology,ound that the price of coffee, he keyEthiopian xport,had a significant ositive mpacton Ethiopian apitalgood mports nddomestic apitalformation.De Gregorio 1992)analyzedgrowthdeterminantsn twelveLatin American ountriesover the period1950-1985.Controlling or investment nd macroeconomictability,he did not find a significant ffectof the termsof tradeon growth.However, e did not consider hat investment tselfmightbe affected ythe termsof trade.III. The DataWe consider he period 1965-1989.Our sampleconsistsof the following18oil-exporting ountries:Algeria Gabon Mexico Trinidad ndTobagoBahrain Indonesia Nigeria UnitedArabEmiratesCongo Iran Oman VenezuelaEcuador Iraq Saudi ArabiaEgypt Kuwait Syria

For each country,on the expenditure ide we examineddata on consumption nd investment by theprivate ector, y the government,nd n the aggregate),avings, he radebalance, ndGDP. At a sectorallevel, we analyzedvalue added n the oil sector and the non-oilsector. The non-oilsectorwas furtherbrokendown nto he following ategories: griculture,manufacturing, onstruction, ublicutilities,and

7/29/2019 107507322_20041117185522

11/42

page 8services. Serviceswere n tum disaggregatednto transportation communications, holesale& retailtrade, and other services. Tradables re probablymost appropriatelydentifiedwith agriculture ndmanufacturing,ontradables ith he othernon-oilcategories.All he abovevariableswere n per capita,constant ocalcurrency erms. As explanatory ariables,we useddata on the termsof trade,a debtcrisisdummy, nd the world eal interest ate.All variableswereat an annual requency.A fulldescription fthe data used and its sources s in Appendix1.IV. Econometric ramework and issues.

A. EstimationThroughout,we use a panel ier hana time-seriesmethodologyince, f the underlying ssumptionsresatisfied, his allows or morepreciseestimates nd rendersomittedvariablebias a lessseriousproblem(seenextsub-section).Wehaveno strongpriors,anddteoryprovidesittle guidance n thespeedat whichour variables djust to terms-of-tradehocks. Thereforewe take whatwe think s a flexibleapproach othe issueof dynamics.As a first pass, we try to pick out long-ran elationships y selectingyearswhensufficient time has passed under a given regime for the countries o be in long-runequilibrium.Specifically, e note hat 1973was he endof a longperiodof stableoil prices; n 1981, fter eightyearsof risingprices,agentsgeneallyacceptedhe higherpricesas prmanent, and hadhad some ime o adjustto them; n 1989, fter an eight-year eriodof fallingprices,agents ikewiseaccepted nd had had sometime o adjust o the lower rices;and,after 1989,oil prices osr. harply round he time of the Gulf War.Thus in 1973, 1981,and 1989 he countries n our samplecould arguablyhave achieved ong-runequilibrium. We thereforeuse these hreebenchmark ears o cartyout fixed-effects stimation f thefollowing anel regression quation:

y -t at,+ TOT,+ en , (1)

wherey denotes he natural og of eachof the variableswe study(except or the tradebalance,which sleft in levels),TOTdenotes he og of the termsof trade, he i subscript efers o countries, nd t refers o

7/29/2019 107507322_20041117185522

12/42

page 9time. The elasticityof each variable with uspect to t toerms f trde is given by P. The regression esultsare given n Appendix 1, able 1.In addition,we also use the full dataset to carry out fixed-effects stimation f the following egressionequation,whichallows or lagged esponses:

y - g + ITOT1 + P2TOTIr, + yCO) OL , +e(, !2)

whereCONTROL enotes he controlvariables,whichpotentiallynclude he debt crisisdummyand theworld real interestrate. Because here s little theoreticalguidance vailableon whenand whetherweshould nclude hedebt crisisdummy,we excludedt where t was nsignificant. he results or the worldreal interest rate were hard to interpret,and we always excluded t. The only exceptionwere theinvestment quationswhere,given he strong heoretical riors,we alwayscontrolled or the debt crisisand for the world ealinterest ate. The short-run lasticityof each variablewithrespect o the tens oftrade s givenbyJ,,he steady-state lasticity y (P, + N3).The regression esultsare given n AppendixII, table2, and are discussed n the nextsection.Note hatour specification f equations1) and (2)makes e important ssumptions.First, he randomefrors s, avezeromeanand are .i.d.over ime andacrosscountries.Second, he ntercept variesacrosscountries but not over time),so that therearecountry-specific ffects. I subject his hypothesiso thefollowing aange Multiplier est,due o Breuisch& Pagan 1980).Let a be a random erm with cross-country variancea.. Let the null hypothesis e that there are no country-specific ffects, hat is,Ho: a. =0, versusHI: a. > 0. Let e,tdenote he residuals rom the standardOLS regressionwithhomogeneousntercepts.UnderHo, he test statistic

LAM N T I3 d X21).2(T-1) iS

7/29/2019 107507322_20041117185522

13/42

page 10

The asymptotic -values rereported n Appendix . In all regressions, M is significant t the .0001%level; we therefore eject the null. Note that, since the test statistichas a block-diagonalnformationmatrix, hispre-testdoes not affect he standard rrorsof the otherestimatorswecompute.Third, we assume hat the elasticitieswith respect o the termsof trade (thep coefficients) re constantover time and acrosscountries. We est both of thesehypotheses.To investigate emporal tability,wesplit up the sample nto the subperiods1965-1980 nd 1981-1989;ince heyroughly orrespond o theperiodsof rising ndof failing ermsof trade, hisalso allowsus to checkwhether heresponseso positiveand to negative erms-of-trade hocksdiffer. As detailed n Appendix L table 3, the null hypothesis fidenticalb oefficients cross he two subperiods s generally upported y the F-testdescribed n Hsiao(1986),chapter .2,equation2.2.20;where he differences cross ubperiods re statistically ignificant,they are typically uantitatively nimportant.Thissuggests here are no significant symmetriesn theresponse o risingand to falling ermsof uade.In contrast, s detailed n Appendixm, thenull hypothesis f stability crosscountries s always ejectedat the 1% significanceevelby the F-test described n Hsiao 1986),chapter2.2,equation2.2.15. Therejections hardly urprising, iven hat we aredealingwith such a broad sampleof countries, ut it doesimply that we must look at country-specificegressions o determinehow representative he panelestimates re. Thesecountry-by-countryegressions re accordingly resented n Appendixm.Finally,note hat our fixed-effects stimator s onlyBLUE f we interpret he country-specificffects 1as fixed regressors.To forecast n out-of-sampleesponse o a terms-of-trade hock,we musthowevertreat the a%'s s randomerrors;our estimator s then in generalnot BLUE,but is still preferable o therandom-effectsstimatorsince t is consistent ven when the az's are correlatedwith the explanatoryvariables.

B. Omitted ariable biasThe ssueof bias fromomittedvariablesneeds o be treateddifferentlyn apanelcontext. To see his, tis convenient o define he mean across ime for any givencountry nd variable,x, as:

7/29/2019 107507322_20041117185522

14/42

page 11

Xi - (4)and the deviationsfrom country-meansas:

Xs X. -X. (5)

Re-writtenin terms of deviationdata, equation (1) thenbecomes

YU,- pTOT,+ 1 . (6)

Saythat the truemodel is instead given by

.= . PTOT,, yOAM + X - (7)

If equation (5) were estimated by OLS for each country, the bias b; in our estimatorof tating theregressors as fixed, wouldbe

TE T6Tt , y OMEDI(,bf T (8)

E ToT2

Hence, a necessaryandsufficientconditionforthe countrty-by-countyOLSestimatesof i to beconsistentis that asymptotically he termsof trade be orthogonal o any omittedvariables:

Al. lim(T--) b,r = 0.

7/29/2019 107507322_20041117185522

15/42

page 12Now say that equation 5)is estimated n a panelcontext.The bias in ouLrstmatorof P is then

Ir rE 7T&T * y 0MJZTAD,,b; I1 .1 Ir (9)

T 6T

- t*ib-I (10)

Hence, a necessary and sufficientcondition for the panelesfimatesof p to be consistent s thatasymptoticallyweightedaverage f the country-specificiasesbe zero:

A:2. lim(T--) F,,aTUb= 0.Clearly,Al impliesA2 but not hereverse: herearemanyways or assumptionA2 to be satisfiedwithoutanyof the countries ndividuallyatisfying ssumptionAl. The importantpoint is thatfor the panelestimates o be consistent, t is unnecessaryor the termsoftrade o be asymptoticallyrthogonal o theomittedvariablesn anycountry:whatmatters s that he country-specficsymptoticiasesbe negativelycorrelatedacrosscountries o that theyaverage o zero. Thislatter assumptions morecredible o theextent hat the relationship etween he omittedvariablesand he termsof trade s idiosyncratico eachcountryandnot positively orrelated crosscountries.Of course, o theextent hat he omittedvariables re worldvariableswhichaffect all countries, hen heb; ermswill tend to have the same igns and assumptionA2 mayfail. We therefore ontrol or thedebtcrisisandfor world nterst rates. Etmightalsobe desirable o controlor worldoutput,but sinceperiodsof rising oil pricescoincidedwith changesn OECDactivity hatwouldhave depressed rowth n oursample, ndvice-versa,hiswouldprobably trengthenhe linkswe find between erms-of-tradehangesand activityvariables.

7/29/2019 107507322_20041117185522

16/42

page 13The additional ssue of the weights, a,, is unlikely to be important for our sample. Countries have largerthan average weights to the extent that the sample variability of their terms-of-tradedata is larger thanaverage. Since our sample consists of oil exporters, the variability of their terms of trade is similar.

C. Non-stationarityThere are two ways to inltpret our econometricmodel. Frst, we may view the terms of trade (which aredriven mainly by the oil price) as a fixed regressor. Equations (1) and (2) are then the mechanismgenerating he rest of the data. The presence or absence of stationarity is not an issue in this framework.Altemvatively, e may treat the terms of trade as being generated by some time-series statistical model.Here, the order of integrationof our data is an issue. Unfortunately, ittle is known about how to test forand deal with nonsaionarity in a panel data setting. When estimating equation (2) for each country inisolation,we carry out an EngleGranger co-inwgration est, using the augmentedDickey-Fuller statistics.For most countries and many variables, we cannot reject the null hypothesis that the regression residualsare integrated of order 1. However, given that such tests are well known for their very low power, andgiven the paucity of observationsat the single-country evel, we do not feel there is much strong evidencethat our data is non-stationary n the sense that it trends over time.V. Results

A. The importance of InvestmentOverall,we find fairly strong evidence hat the relationshipbetween erms-of-trade hocks and investmentis positive. The panel regression suggests the 95% confidence nterval (0.500,0.675) for the long-runelasticity. In countryby country egressions, 14 of the 18 countrieshave positiveelasticities and of the 14,13 are stastically significant This section examines what the availableevidence says about the sectoralbreakdown and the causes of this investment effect, and how this investment effect is crucial tounderstandingother effects such as those on the trade balance and on growth.

7/29/2019 107507322_20041117185522

17/42

page 14It is likely that much of this investmnentccurred n the nontradable ector of the economy. Although dataon investmentby sector s simply not available or mostof these countries,' and therefore his claim cannotbe supported directly,much of the otber data we do have is supportive. First, there s strong evidence thatthe relativepriceof nontradableswere positivelyassociatedwith the terms-of-trade hocks. So there seemsto have been a clear price incentive for investment n these sectors. Second, we do have evidence that inthe long run output expanded in nontradable sectors such as construction and transportation. It seemsreasonable to think that this rise was made possible by a higher capital stock in these sectors.This investmenteffi- in nontradablescan be explained ormally in a variety of intertemporaloptimizingmodels of investment (recent examples in the literue include Schmidt-Hebbel& Serven (1993) andWarner (forthcoming, 1994)). These models usually conclude that aggregate private investment isdetermined by the ratio of Tobin's q (the present discounted value of future marginal value products ofcapital in a given sector) to the relevant price deflator for investment goods. Since the countries in oursample import manycapital goods, and the world pricesof capital goods did not change dramatically overthe period, the price of nontradables relative to physicalcapital probably rose in the 1970's and declinedin the 1980's,causing similar changes in investment n the nontradable sector.Tuming to the oil sector,one might think that relativeprice changes would also stimulate nvestment here.Although his would probably hold for countries experiencingexogenous changes n their export prces,it is important to remember hat OPEC raised oil prices n the 1970's partly through a deliberate policy ofoutput restriction,so that for much of our period the oil sector was engineering price changes rather thanpassively responding o them. In such a setting, where countries exercise their power in the world oilmarket, it is a prioriunclear which way the incentiveswork for investment n the oil sector. Oil productionquotas may be seen as a permanent ceiling on oil production, leaving little reason to invest. On the otherband, an increase in productive capacity may strengthen one's hand in the periodic quota negotiations.

XAn exceptions Wamer 1994),where t is shown hat nvestmentn Mexican rivate on-tradablesid declineover 1981-85.That tudy lso suggests hat an importantmechanismor the investment eclinewas the fall in theratioof nontradablences o imported quipment rices, nd hat the erms-of-tradeecline anexplainmostof thereduction n this relative rice.

7/29/2019 107507322_20041117185522

18/42

page 15Therefore, we cannot deduce from theoly whether investment should have risen or declined in the oilsector.We do know hat real oil output per capita did not rise over the long run between 1973and 1981. It maybe thatcountries investedin modernizing he oil sector rather thanin expanding its capacity. It may alsobe that they did invest in expandingcapacity but just did not use this capacity fully over the period. Wesimply do not have the data to pin this down very precisely.However, it is clear that overall investment responded positively to the terms of trade, and this helpsexplainwhy the trade balance was strongly negatively related toterms-of-trade shocks, even though theexport price effect in isolation should cause a higher trade surplus.

B. Consumption, Savings,and the Current AccountAs discussed above, tory suggeststhat after a permanent positive terms-of-tradeshock, consumptionshould rise, and savings should as a first approximation be unaffected. The data broadly confirmstheseimplications. The only exceptionswere Ecuador,Egypt, Syria, Kuwait, and the Emirates, which displayedan insignificant or indeed negative consumption response to the terms-of-trade shock. However, theformer hree countries were not really oil exporters before the firstoil price shock, which consequentlyhada smaDler ositive impact on their wealth. For Kuwait and the Emirates, he shock seems to have induceda large ncrease in the consumptionof leisure, made up for by extensive immigration which pushed downaverage evels of consumptionpercapita. The data also suggestsgovernment consumptionmay respondalmost twice as strongly as private sectorconsumption,perhaps because govemmentssystematicallyviewshocks as more permanent than is the case. Numerically,panel estimates suggest the 95% confidenceinterval (0.310, 0.426) for the elasticity of consumption with respect to the terms of trade.Given the lack of an impact on savings, and the strong response of investment, the trade balance andcurrent account should be expected o decrease,and without exception these were the estimated esponses.

C. Nontradable Prices and QuantitiesTerms-of-trade hocks are verystrongly associated with real exchange rate appreciation, i.e., increasesinthe relativeprice of nontradables. Panel estimates of the elasticity suggest the 95% confidence interval(-

7/29/2019 107507322_20041117185522

19/42

page 160.606,-0.695), ndonly Congodisplayed n insignificantesponse. The shocks re also associatedwithsignificantncrease n valueaddedwithineachcategoryof nontradables.This confirms he importanceof spending ffectsas a key mechanism n the transmission f terms-of-tradehocks o the rest of theeconomy: igherwealth eads o greater emand or nontradables,ndhence o an increase n their relativepricesandquantities.At a country evel, he mainexceptionwas Ecuador,whichdisplayed nsignificantresponses n the value addedof each nontradable ector.

D. DutchDiseaseEffectsThe iterature trongly ocuseson DutchDisease ffects;here, the two non-oil radables agriculare andmanufacturing)re the closestapproximation e have to a DutchDisease ector. Yetwe failed o detectclearcontractionsn eitherof them n responseo a rise in the priceof oil; he only exceptionwas Nigerianagriculture.Oil valueaddeddidrespond egatively ver hesubperiod1965-198, but this simply eflectsthe impositionof OPEC quotas: he oil sector s best not viewedas facingexogenouslygiven ermsoftrade,sincepart of OPEC's strategy t the time clearlywas o engineer ricerises by restraining utput.The absence f DutchDiseaseeffectsmaybe partlyexplainedby either he compression n the outputofthe oil sector,or by its typically eingan 'enclave' sector whichdoes not participate n domestic actormarkets.Both actorswouldact to reduce he pull of resourcesaway romothersectors,as mentioned nthe theoreticalection.Nevertheless, s discussed bove, he spending ffectwasoperational, ncreasingrelativeprices and output n the nontradable ector. To the extent hat the nontradable ector maybecompetingwith agricultureand manufacturingor scarce actorsof production, r if nontradables rethemselves ntermediatenputs into other sectors,one might have expecteda small bout of the DutchDisease.

E. GrowthGiven hat herelativepricesbetween il andotherproducts hanged o dramatically ver he period,andthat theoil sector s so important n our sample,extremely erious ndexnumberproblemsarise n evendefiningand measung aggregateoutput. We therefore feel that it is not analytically useful to talk aboutthe aggregate conomy, nd we instead eparate he oil from he non-oil ector. The formerhas alreadybeen mentioned. n the non-oilaggregate,we do see an effect on GDP,drivenmainlyby the expansion

7/29/2019 107507322_20041117185522

20/42

page 17in nontradables;Ecuadorand Nigeriawere the only countrieswith insignificant esponses. Givenourevidence n investment,his response s best seen as beingdrivenby capital ormation n the nontradablesector. Numerically, anel esimates of the elasticityof non-oilvalue addedwith respect o the termsoftrade suggest he 95% confidence nterval 0.206,0.297). This is however nfluenced y outliers; hemedianelasticity rom the country egressionss 0.381.VI. Some Warnings regarding ExtrapolationWhen interpreting nd aboveaUextrapolating ur results,several ssuesmustbe considered.First, hecauses of the oil price shock were different rom the causes of the forecast uture increase in primarycommodityprices. IJ our sample export prices rose because of oligopolistic oordinationaimed atreducing utput. Some esponsesmight wellbe different or a truly exogenouserms-of-trade hock. nparticular,both investment nd output n the favoredsector wouldprobably ise; since output was nolongerconstrained,we shouldalso expecta bigger ncrease n wealthandhenceconsumption.Second,our resultsneedonlybe valid or a permanent hange n the termsof trade. A temporary erms-of-tradeshockmay lead to some ntertemporal ubstitution n consumption nd nvestment, ut is unlikelyto causea permanent hange n output.Third,an economy's esponse o an increasen thepriceof primary ommodities ill clearly ingeon howdependenttis on primaryexports. Its investment esponsewillalsodependon what shareof its capitalgoods s imported.Fourth,we mightexpect he availability f external inance o be a crucial actor nteractingwith changesin the terms of trade,and oil exporters n the 1970'sprobably njoyedbetter access o the world capitalmarkets an is trueof the developing orld oday. On the otherhand, his may wellchange f the trendsin commodity riceschange.Fifth, in the 1970'sand 1980'soil-exporterswasteda lot of their windfall evenueson prestigeprojectswith little impacton output. Presumably, eveloping ountrieshaveby now earnt he lesson,and nexttime aroundwill makebetter use of their luck.

7/29/2019 107507322_20041117185522

21/42

page 18VII. ConclusionsWeconsidered panelof oil-exportershat mporta significantractionof their capitalgoods. We foundthatpermanenterms-of-tradehockshaveno impacton savings, strongpositive mpacton investment,and a negative mpacton the currentaccount. There is also evidenceof a long-runeffecton output,particularly f non-tradables.Real exchange-rate ppreciationsre a key mechanism n triggering heresource eallocation. We failed o find any evidence hat the Dutchdisease is a majorphenomenon.Finally, the responseof expenditure o terms-of-trade hocks is not very sensitive o whether heexpenditure omes rom hepublicor privatesector.As discussedn sectionVLestimating n in-sample esponseo a terms-of-tradehock s much asier hanforecastingn out-of-sampleesponse.Whileour qualitative esultsmayhold in quitegeneral ontexts,our quantitative stimates lmostcertainlywill not. Let us neverthelessry to obtainsome dea of thepotentialmpactof theforecast ncreasen commodity riceson the growth ateof commodity xporters.As a lower-boundcenario, ssumehat hecommodityectordoesnot growat all, hat henon-commoditysectoraccounts or 75% of GDP, nd hat it responds ike he non-oil ector n our sample. An increasein the trendgrowth ateof commodityrices rom-1.5%p.a to 0.7%p.a. thenimplies n increase n theorderof 0.6% n the trendgrowth ateof per capitaGDPat constant rices. To he extent hatcommodityoutput ncreases, nd hat hisdoesnot pull esources way rom he restof the economy,he growth ffectis magnified.

7/29/2019 107507322_20041117185522

22/42

page 19Appendix I: Full Data Description and Sources.We consider the period 1965-1989. Our sample consists of the 18 oil-exporters isted below.

Algeria Gabon Mexico Trinidad and TobagoBahrain Indonesia Nigeria United Arab EmiratesCongo hran Oman VenezuelaEcuador Lrq Saudi ArabiaEgypt Kuwait Syria

For all countries, fuels accounted for over 50% of total exports over at least half the sample period; formost countries, fuels accounted for over 70% of total exports over at least three quarters of the sampleperiod. We used data on the following variables:(1) Termsof Trade (TOT), computed as Merchandise exports deflator US Dollar (USD)I merchandise

imnports eflator USD, base 1987= 1.0.(2) A Debt Crisis dummy (DEBT),set to unity for Ecuador,Mexico,Nigeria and Venezuelaover 1982-1989, and to zero elsewhere.(3) World Real Interest Rates (INTRATE), s computed by Barro and Sala-i-Martin.(4) The Real Exchange Rate (REALEXRA), or relative price of tradables to nontradables. This iscomputed as USA GDP derlator USDI local GDP deflator USD. index 1987.(5) Gross Domestic Product (GDP),per capita, constant 1987marketprices,Local CurrencyUnit (LCU).(6) Gross National Product (GNP), per capita, constant 1987 marketprices, LCU.(7) Consumption (CONS), per capita, constant 1987 prices, LCU.(8) Private Consumption (CONSPRIV),per capita, constant 1987prices, LCU.(9) General GovernmentConsumption(CONSGOVT), per capita, constant 1987 prices, LCU.(10) Gross Domestic Savings (SAVINGS),per capita, constant 1987prices, LCU.(11) Fixed Investment (INVT), per capita, constant 1987prices, LCU.(12) Private fixed Investment (INVTPRIV),per capita, constant 1987prices, LCU.(13) General Government fixed Investment (INVTGOVT),per capita, constant 1987 prices, LCU.(14) Balance of merchandise Trade (TRADBALA), per capita, constant 1987 prices, LCU.(15) Value added in the Oil sector, constant prices (OILVA),LCU.(16) Value added in the Non-oil sector (NONOILVA), onstant prices, LCU.(17) Value added in Agriculture (AGRICULT),constant prices, LCU.

7/29/2019 107507322_20041117185522

23/42

page 20(18) Valueadded in Manufacturing (MANUFACT), onstant prices, LCU.(19) Valueadded in Construction (CONSTRUC),constant prices, LCU.(20) Valueadded in Public Utilities (UTILITY),constant prices, LCU.(21) Value added in Services (SERVICES),constant prices, LCU.(22) Value added in Transportation and Communications TRANSPORT),constant prices, LCU.(23) Valueadded in Wholesale and Retail Trade (TRADE), constant prices, LCU.(24) Valueadded in Other Services (OTHRSERV), onstant prices, LCU.(25) Total Labor force (LABOR).

Variables 1, 4.. 14 and 25 came from the World Ban's DAD database (except for Mexico's terms oftrade, obtained from Mexico's central bank), and variables 2 and 3 are as described above, They areavailable for all countries for the whole period.Variables 17, 18 and 21 were from the WorldBank's STARS database, and all other variables came fromWorld Bank CountryEconomic Memoranda. They were not available for Iraq, nor before 1969.In all cases, the ultimate sources are national central banks, national statistical services, and estimates byWorld Bank missions.

7/29/2019 107507322_20041117185522

24/42

page 21Appendix I: Full Panel Regression Results.Table 1. Panel regression: In yi, = a, + 1BnTOTi, + e, . Sample period: 1973, 1981, 1989.

Dependent variable Terms of trade Terms of trade T- NNCoefficient ratioLn Real exchange rate -0.5617 -6.8338 0.5716 0.3973 54Ln Consumption 0.3594 4.4355 0.3598 0.3917 54Ln Private Consumption 0.3413 4.2327 0.3386 0.3898 54Ln Govt Consumption 0.4085 4.1049 0.3250 0.4810 54Ln Savings 0.0312 0.2177 0.0014 0.6939 54Ln Investment 0.4780 4.3860 0.3547 0.5267 54Ln Private Investment 0.4895 3.5540 0.2652 0.6656 54Ln Govt Investment 0.5731 4.5936 0.3761 0.6030 54Trade Balance -15.1420 -2.8571 0.1936 24.9415 53Ln GDP -0.0123 -0.1613 0.0007 0.3689 54Ln Oil value added -0.3862 -1.7479 0.2763 0.9133 23Ln Non-oilvalue added 0.2211 4.4454 0.7118 0.2055 23Ln Agriculture 0.0706 0.7940 0.0220 0.3787 46Ln Manufacturing 0.1710 1.2890 0.0767 0.5432 36Ln Construction 0.3314 2.7997 0.4655 0.4852 25Ln Public Utilities 0.2553 1.5724 0.2610 0.6039 21Ln Services 0.0951 1.0117 0.0353 0.4003 46Ln Transportation 0.1958 1.7262 0.2986 0.4323 22Ln Trade 0.1419 0.8500 0.0936 0.5207 21Ln Other services 0.2704 3.7993 0.6734 0.2713 22

7/29/2019 107507322_20041117185522

25/42

page 22Table 2. Panel regression: In Yi = cc + 3,BnTOTi, + 12 In TOTi,.I + .. + E,.

Dependentvariable Terms of trade sum Tenns of trade T- R2I Nof coefficients ratioLn Real exchangerate -0.6508 -28.7336 0.6747 0.2488 431Ln Consumption 0.3681 12.5055 0.2766 0.3248 431Ln Private Consumption 0.2902 10.2829 0.2105 0.3114 431Ln Govt Consumption 0.5206 12.9296 0.2892 0.4442 431Ln Savings -0.0679 -1.1228 0.0636 0.6660 429Ln Investment 0.5875 13.1991 0.3217 0.4706 413Ln Private Investment 0.6831 11.7508 0.2902 0.6146 413Ln Govt Investment 0.5970 12.3281 0.2874 0.5120 413Trade Balance -15.4053 -9.2591 0.2145 15.9832 395Ln GDP -0.0201 -0.7378 0.0158 0.3000 431Ln Oil value added -0.3640 -6.3850 0.2470 0.3548 178Ln Non-oil valueadded 0.2519 10.8723 0.4316 0.1441 178Ln Agriculture 0.0701 2.2620 0.0691 0.2155 293Ln Manufacturing 0.2306 3.7061 0.0803 0.4022 240Ln Construction 0.3967 9.1806 0.3581 0.2709 193Ln Public Utilities 0.3163 6.3042 0.3226 0.2787 158Ln Services 0.0965 2.7620 0.0276 0.2438 293Ln Transportation 0.2332 5.4244 0.1758 0.2467 175Ln Trade 0.2331 4.6300 0.1578 0.2382 158Ln Other services 0.2606 9.7659 0.3833 0.1531 175For the real exchangerate, the trade balance, non-oil value added, agriculture,construction, utilities, andtrade, the regressor matrix also includes DEBTj,; n all the investment equations, it also includes DEBT1and In LNTRATE,. Sample period: 1965-1989.

7/29/2019 107507322_20041117185522

26/42

page 23Table 3. Panel Regression: In y,1 = oc + 01 In TOT,, + 132In TOT,,., + .. + ell. coeficientsunrestrictedacross the subperiods 1965-1980 nd 1981-1989.

Dependent variable Terms of trade Terms of trade P-value for R2 Nsum of coef. sum of coef. coef.1965-1980 1981-1989 stabilityLn Real exchange rate -0.6746 -0.5484 0.0235 0.6807 0.2471 431Ln Consumption 0.3791 0.3110 0.25 0.2815 0.3245 431La PrivateConsumption 0.3042 0.2276 0.403 0.2140 0.3115 431Ln Govt Consumption 0.5395 0.4211 0.0931 0.2974 0.4428 431Ln Savings -0.1439 0.1884 0.0894 0.0747 0.6637 429Ln Investment 0.3124 0.3510 0 0.4505 0.4253 413Ln Private Investment 0.4647 0.1246 0 0.3689 0.5819 413Ln Govt Investment 0.2590 0.6564 0 0.4226 0.4627 413Trade Balance -16.9939 -10.9972 0.419 0.2181 15.9906 395Ln GDP -0.0488 0.0699 0.0893 0.0274 0.2990 431Ln Oil value added -0.4854 -0.0789 0.0258 0.2807 0.3491 178Ln Non-oil value added 0.2449 0.2710 0.703 0.4341 0.1448 178Ln Agriculture 0.1125 0.0087 0.259 0.0784 0.2153 293Ln Manufacturng 0.2021 0.2877 0.281 0.0908 0.4018 240Ln Construction 0.3898 0.4083 0.76 0.3601 0.2721 193Ln Public Utilities 0.3213 0.3142 0.713 0.3258 0.2802 158Ln Services 0.0842 0.1241 0.399 0.0341 0.2439 293Ln Transportation 0.1601 0.4157 0.0798 0.2020 0.2445 175Ln Trade 0.2396 0.2250 0.936 0.1585 0.2400 158Ln Otherservices 0.2597 0.2674 0.53 0.3883 0.1536 175For the real exchange ate, the trade balance,non-oil value added, agriculture,construction, utilities, andtrade, the regressor matrix also includesDEBTk; n all the investmentequations, it also includes DEBT,,and In INTRATE. Sample period: 1965-1980, and 1981-1989. The P-value denotes the minimumsignificance evel at which we can reject the null hypothesisof identical erms-of-trade oefficients acrosssubperiods,using the F-test described n Hsiao (1986), ch. 2.2, equation 2.2.20.

7/29/2019 107507322_20041117185522

27/42

page 24Table . Share f capital oods n total nports n 1973, n 1981, nd n he latest vaiableyearup o 1989inclusive.

Shareof capital oods n theyear Latest vailable ear-Country 1973 1981 Latest 1989 f not statedAlgeria 0.3720 0.3834 0.2570Congo 0.3868 0.3874 0.3620 1986Ecuador 0.4117 0.4582 0.3397Egypt 0.2477 0.2816 0.2303Gabon 0.4066 0.4094 0.4009 1983Indonesia 0.4104 0.3536 0.3774Iran 0.3779 0.2832 0.3884 1983Iraq 0.3297 0.5328 0.4478 1983Kuwait 0.3442 0.4103 0.2961Mexico 0.4438 0.4707 0.3260Nigeria 0.4012 0.4402 0.3812 1986Oman 0.3101 0.3908 0.3685SaudiArabia 0.3524 0.4047 0.3870Syria 0.2369 0.2171 0.2592 1986Trinidad ndTobago 0.1313 0.2242 0.2729UnitedArabEmirates 0.3801 0.3568 0.3025 1986Venezuela OA688 0.4344 0.4682 1988

7/29/2019 107507322_20041117185522

28/42

page 25Appendix II: Country-by-countryRegressionResults.Sampleperiod: 1965-1989. or each dependent ariable, wo regressions recarriedout.Regression : In yj,= uxi P(L) nxk, e,1, j identical crosscountries.Resultsare in the form:Y variable TOTCoef TOTT-stat DebtCoef DebtT-stat IntRateCoef IntRateT-statR2 a N P-valueRegression : In yt = a;+ 3P(L) Inxt + ej,, B nrestrctedacrosscountries.Resultsare in the form:Country TOTCoef TOTT-stat DebtCoef DebtT-stat IntRateCoef IntRateT-statR2 a N P-valueThe P-value n regression refers o the LagangeMultiplier est of thenull hypothesis f no country-specific ffects;heP-value n regression to theF-testofthe nullhypothesisf coefficient tabilitycrosscountries. Wherea coefficient nd its T-statistic re not reported, he corresponding ariablewas notincluded n the regression.REALEXRA -0.651 -28.7 0.325 6.070.675 0.249 431 4.21e-17Algeria -0.447 -6.82Bahrain -2.53 -13.6Congo -0.0493 -0.257Ecuador -0.263 -1.89 0.166 1.87Egypt -0.818 -3.69Gabon -0.629 -6.29Indonesia -0.467 -5.25hran -0.691 -11.5Iraq -0.866 -12.3Kuwait -0.728 -13.4Mexico -0.637 -2.29 0.178 1.53Nigeria -0.731 -8 0.489 4.99Oman -0.853 -15.5SaudiArabia -0.783 -14.8Syria -0.214 -1.55Tobago -1.14 -5Emirates -0.577 -9.97Venezuela -0.359 -5.48 0.354 3.280.804 0.203 431 2.7e-14CONS 0.368 12.50.277 0.325 431 5.81e-07Algeria 0.488 6.03

7/29/2019 107507322_20041117185522

29/42

page 26Bahrain 0.917 4Congo 0.145 0.613Ecuador -0.261 -1.57Egypt -0.64 -2.34Gabon 0.682 5.54Indonesia 0.596 5.45Iran 0.292 3.95Iraq 0.186 2.15Kuwait -0.00163 -0.0244Mexico 0.276 0.993Nigeria 0.327 3.25Oman 0.671 9.9SaudiArabia 1.09 16.7Syria -0.513 -3Tobago 0.794 2.84Emirates -0.0429 -0.603Venezuela 0.225 3.320.609 0.25 431 0CONSPRIV 0.29 10.30.21 0.311 431 1.94e-10Algeria 0.48 6.25Bahrain 0.538 2.47Congo 0.202 0.898Ecuador -0.23 -1.45Egypt -0.666 -2.56Gabon 0.539 4.6Indonesia 0.544 5.23Iran 0.256 3.64Iraq 0.0561 0.682Kuwait 0.00356 0.056Mexico 0.297 1.12Nigeria 0.312 3.27Oman 0.425 6.59SaudiArabia 0.99 16Syria -0.502 -3.09Tobago 0.856 3.21Emirates -0.232 -3.42Venezuela 0.23 3.570.581 0.237 431 0CONSGOVT 0.521 12.90.289 0.444 431 3.82e-08

7/29/2019 107507322_20041117185522

30/42

page 27

Algeria 0.51 4.34Bahrain 2.21 6.62Conglo -0.0562 -0.163Ecuador -0.454 -1.88Egypt -0.501

-1.26Gabon 1.03 5.74Indonesia 0.995 6.25Iran 0.418 3.89Iraq 0.371 2.95Kuwait -0.0119 -0.123Mexico 0.106 0.263Nigeria 0.414 2.83Oman 1.03 10.5SaudiArabia 1.23 13Syria -0.557 -2.25

Tobago 0.617 1.52Emnirates 0.329 3.18Venezuela 0.195 1.970.567 0.363 431 0

SAVINGS -0.0679 -1.120.0636 0.666 429 9.34e-16Algeria 0.462 2.37Babrain 2.76 4.5Congo 0.015 o.0264Ecuador -1.14

-2.83Egypt -0.212 -0.322Gabon 0.787 2.65Indonesia 1.4 5.31Iran 0.0746 0.419Iraq -0.258 -1.24Kuwait -0.76 -4.72Mexico 0.256 0.383Nigeria -0.186 -0.768Oman 0.232 1.42SaudiArabia -0.288 -1.82Syria -0.931 -2.27Tobago 1.09 1.62Enmirates -0.345 -2.01

Venezuela -0.271 -1.66 9.18e-110.303 0.601 429 --1

INVT 0.587 13.2 -0.368-3.34 -o.00073 -11

7/29/2019 107507322_20041117185522

31/42

page 280.322 0.471 413 4.16e-10Algeria 0.743 6.04 -0.00298 -1.42Bahrain 1.76 5.09 -0.00417 -1.99Congo 0.282 0.618 0.00402 1.61Ecuador -o.246 -0.798 0.0597 0.248 -0.00117 -0.39Egypt -1.3 -2.81 0.00772 3.83Gabon 1.41 7.3 -0.00653 -3Indonesia 1.41 8.5 0.00233 1.13Iran 0.401 3.57 -0.00636 -3.04Iaq 1.03 7.63 -0.00198 -0.955Kuwait 0.316 3.19 -0.00202 -0.975Mexico 1.01 1.89 -0.0707 -0.264 o.oD0483 0.19Nigeria 0.624 3.67 -0.389 -1.78 -0.00697 -2.69Oman 0.68 6.84 -0.00163 -0.785Saudi Arabia 1.33 13.7 -0.000509 -0.245Syria -0.554 -2.09 0.00413 1.98

Tobago 1.77 3.87 -0.000137 -0.0675Emirates -0.0034 -0.0322 -0.00318 -1.53Venezuela 0.25 2.11 -0.487 -1.93 -0.00265 -1.050.661 0.359 413 3.61e-27i'VTPlRV 0.683 11.8 -0.675 -4.7 8.03e-05 0.09380.29 0.615 413 3.Ole-10Algeria 0.756 5.12 -0.00581 -2.31Bahrain 1.74 4.2 -0.00449 -1.78Congo 1.08 1.97 -0.000525 -0.175Ecuador -0.191 -0.516 0.0579 0.2 -0.00218 -0.603Egypt -0.749 -1.35 0.00223 0.92G3abon 1.29 5.57 -0.00215 -0.825Indonesia 1.36 6.82 0.00169 0.683Inn 0.524 3.89 -0.00478 -1.9Irq 1.62 9.97 0.00611 2.45Kuwait 0.0359 0.301 -0.000757 -0.305Mexico 0.448 0.702 -0.141 -0.439 0.00053 0.173Nigeria 0.495 2.43 -0.635 -2.42 -0.0115 -3.71Oman 1.15 9.62 0.0094 3.77Saudi Arabia 1.74 14.9 -0.00121 -0.486Syria -0.54 -1.69 0.00603 2.41Tobago 1.54 2.8 0.00132 0.545Emirates -0.0215 -0.169 -0.00389 -1.56Venezuela 0.00987 0.0694 -0.488 -1.61 -0.0044 -1.460.701 0.432 413 7.07e-38

7/29/2019 107507322_20041117185522

32/42

page 29lNVTGOVT 0.597 12.3 -0.173 -1.44 -0.00084 -1.180.287 0.512 413 6.48e-15Algeria 0.737 5.3 0.00205 0.865Babrain 1.8 4.61 -0.0035 -1.48Congo -0.747 -1.45 0.0096 3A1Ecuador -0.332 -0.953 0.0624 0.229 0.000259 0.0761Egypt -1.88 -3.59 0.0115 5.03Gabon 1.51 6.9 -0.0139 -5.68Indonesia 1.5 7.98 0.00315 1.35Iran 0.289 2.27 -0.00791 -3.34Iraq 0.973 6.36 -0.00273 -1.16Kuwait 0.444 3.96 -0.00245 -1.05Mexico 2.05 3.41 0.072 0.238 0.000309 0.107Nigeria 0.699 3.64 -0.266 -1.08 -0.00326 -1.11Oman 0.583 5.18 -0.004 -1.7Saudi Arabia 1.1 9.97 -0.000543 -0.231Syria -0.537 -1.79 0.00311 132Tobago 2.06 3.98 -0.00255 -1.12Emirates 0.00568 0.0475 -0.00271 -1.15Venezuela 0.606 4.52 -0.471 -1.65 -0.00213 -0.7470.616 0.406 413 4.86e-14TRADBALA -15.4 -9.26 -11 -3.170.214 16 395 4.07e-13Algeria -2.71 -1.21Bahrain -2.55 -0.362Congo -1.92 -0.267Ecuador -3.2 -0.666 -0.246 -0.081Egypt -0.615 -0.0809Gabon -5.58 -1.64Indonesia -85.3 -28.1Iran -2.31 -1.03Iraq -1.26 -0.506Kuwait -1.37 -0.582Mexico -1.96 -0.207 0.179 0.0449Nigeria -5.06 -1.62 1.09 0.323Oman -4.06 -0.781SaudiArabia -0.522 -0.277Syria -0.784 -0.138Tobago -9.82 -1.27Emirates -1.73 -0.676Venezuela -77.1 -32.2 -1.03 -0.2810.868 6.93 395 3e-108

7/29/2019 107507322_20041117185522

33/42

page 30GDP -0.0201 -0.7380.0158 0.3 431Algeria 0.249 3.18Bahrain 0.051 0.231Congo -0.065

-0.284Ecuador -0.478 -2.97Egypt -0.776 -2.94Gabon 0.607 5.1Indonesia 0.516 4.88Iran 0.0491 0.687Iraq -0.198 -2.38Kuwait -0.589 -9.13Mexico 0.261 0.972Nigeria -0.0206 -0.212Oman 0.352 5.37SaudiArabia 0.17 2.69Syria -0.35 -2.12Tobago 0.69 2.55Emirates -o.324 -4.71VJenezuela -0.142 -2.170.419 0.241 431 0GNP 0.0163 0.6210.0112 0.289 431 2.04e-14Algeria 0.255 3.34Bahrain -0.189

-0.871Congo 0.00967 0.0432Ecuador -. 419 -2.66Egypt -0.753 -2.92Gabon 0.589 5.06Indonesia 0.495 4.79hran 0.0917 1.31Iraq -0.146 -1.78Kuwait -0.472 -7.48Mexico 0.274 1.04Nigeria 0.000775 0.00816Oman 0325 5.07SaudiArabia 0.288 4.68Syria -0.336 -2.09Tobago 0.797 3.01Emirates -0.313 4.66Venezuela -0.126 -1.960.399 0.236 431 0

7/29/2019 107507322_20041117185522

34/42

page 31OILVA -0.364 -6.390.247 0.355 178 4.32e-43Algeria -0.122 -0.916Bahrain 1.1 1.51Congo 1.09 1.01Ecuador -0.138 -0.398Egypt 1.94 2.38Indonesia 0.00744 0.0191Iran -0.776 -6.38Kuwait -0.537 -2.16Mexico -0.173 -0.46Nigena 0.187 0.524Oman -0.202 -0.326SaudiArabia -0.198 -2.01Tobago -1.16 -1.04Venezuela -0.559 -4.93

0.462 0.33 178 0.00347NONOILVA 0.252 10.9 -0.0682 -1.720.432 0.144 178 8.75e-44Algeria 0.336 7.39Bahrain 0.235 0.952Congo 0.92 2.51Ecuador -0.0364 -0.355 0.137 2.32Egypt 0.754 2.71Indonesia 0.591 3.74Iran 0.28 6.75Kuwait 0.091 1.07Mexico 0.417 2.48 0.102 1.48Nigeria -0.00423 -0.0332 -0.117 -0.938Oman 0.847 4.01Saudi Arabia 0.381 11.3Tobago 0.836 2.21Venezuela 0.0778 1.72 -0.246 -3.290.721 0.113 178 3.54e-10AGRICULT 0.0701 2.26 -0.167 -3.39

0.0691 0.216 293 2.33e-18Algeria 0.183 2.08Babrain 0.607 1.36Congo 0.00972 0.0466Ecuador -0.108 -0.631 -0.00651 -0.0722Egypt 0.001 0.00387

7/29/2019 107507322_20041117185522

35/42

page32Gabon -0.0547 -0.208Indonesia 0.151 1.29Ian 0.0964 1.18Kuwait 0.17 2.21Mexico 0.228 0.744 -0.079 -0.601Nigeria -0.193 -2.11 -0.216 -2.06Oman -0.203 -1.25SaudiArabia 0.0405 0.533Syria 0.63 2.41Tobago 0.955 4.13Emirates 0.14 0.709Venezuela -0.0647 -0.756 -0.081 -0.7680.304 0.2 293 0.000138MANUFACT 0.231 3.710.0803 0.402 240 3.25e-28Algeria 0.298 1.84Bahrain -0.227 -0.239Congo -0.0879 -0.188Ecuador 0.232 0.737Gabon 0.278 0.578Indonesia 0.918 4.26Iran 0.199 1.33Kuwait 0.233 0.73Mexico 0.193 0.468Nigeria 0.188 1.2Oman -0.571 -1.92SaudiArabia 0.102 0.733Tobago 0.922 2.18Emirates 1.53 4.24Venezuela 0.0121 0.08330.332 0.368 240 5.92e-05CONSTRUC 0.397 9.18 -0.306 -4.190.358 0.271 193 1.44e-2BAlgeria 0.714 9.07Babrain 0.382 0.893Congo 2.4 3.78Ecuador 0.00434 0.026 -0.0965 -1.03Egypt 0.945 1.96Indonesia 0.707 3.08Iran 0.053 0.739Kuwait 0.133 0.906

7/29/2019 107507322_20041117185522

36/42

page 33Mexico 0.475 1.63 0.0674 0.567Nigeria 0.64 2.91 -0.488 -2.26Oman 1 2.75SaudiArabia 0.704 12.1Syria 0.79 1.96Tobago 1.9 3.93Venezuela 0.232 2.95 -0.825 -6.370.732 0.195 193 2.5e-15UTILIIY 0.316 6.3 0.259 3.210.323 0.279 158 1.85e-50Bahrain 0.074 0.131Congo 0.235 0.28Ecuador -0.271 -1.22 0.603 4.86Egypt 1.02 1.6Indonesia 0.492 1.62Iran 0.354 3.73Kuwait 0.138 0.709Mexico 1.56 2.2 0.429 2.46Nigeria 0.491 1.68 0.0379 0.133Oman 1.25 2.59SaudiArabia 0.406 4.94Tobago 1.17 1.83Venezuela 0.202 1.94 0.11 0.6390.536 0.259 1SS 0.00748SERVICES 0.0965 2.760.0276 0.244 293 1.22e-20Algeria 0.114 1.16Bahrain 0.357 0.713Congo 0.588 2.52Ecuador 0.0865 0.45Egypt -0.449 -1.55Gabon 0.277 0.942Indonesia 0.497 3.76Iran 0.0653 0.713Kuwait -0.296 -3.43Mexico 0.17 0.676Nigeria 0.316 3.31Oman 0.0162 0.0887Saudi Arabia 0.188 2.2Syria 0.985 3.36Tobago 0.489 1.88

7/29/2019 107507322_20041117185522

37/42

page 34Emirates 0.19 0.859Venezuela -0.0328 -0.3690.277 0.225 293 2.01e-05TRANSPORT 0.233 5.42

0.176 0.247 175 5.47e-52Bahrain 0.962 1.99Congo 1.18 1.64Ecuador -0.00442 -0.0236Egypt 1.63 2.99Indonesia 0.636 2A6hban 0.428 5.28Kuwait 0.106 0.641Mexico 1.58 2.73Nigeria 0.37 1.55Oman 1.46 3.53Saudi Arabia 0.144 2.18Syria 0.507 1.12Tobago 1.25 2.3Venezuela 0.0341 0.4510.46 0.22 175 0.000119TRADE 0.233 4.63 -0.132 -2.050.158 0.238 158 4.43e-47Bahrain 1.1 2.75Congo 1.59 2.69Ecuador -0.0376 -0.242 0.111 1.27Egypt 1.25 2.8Indonesia 0.304 1.42Iran 0.569 8.51Kuwait 0.16 1.17Mexico 0.344 1.27 0.0616 0.556Nigeria 0.00838 0.0409 -0.0632 -0.315Oman 1.04 3.06Syria 0.443 1.18Tobago 0.914 2.03Venezuela -0.0189 -0.26 -0.373 -3.090.61 0.182 158 6.55e-10OTHRSERV 0.261 9.770.383 0.153 175 1.14e-50Bahrain -0.554 -2.02Congo 0.634 1.56

7/29/2019 107507322_20041117185522

38/42

page 35Ecuador 0.0369 0.347Egypt 0.816 2.65Indonesia 0.437 2.98hran 0.305 6.64Kuwait -0.0263 -0.28Mexico 0.389 1.19Nigeria 0.00246 0.0182Oman 0.576 2.46Saudi Arabia 0.42 11.3Syria 0.7 2.72Tobago 0.861 2.78Venezuela 0.0801 1.870.663 0.125 175 1.53e-08

7/29/2019 107507322_20041117185522

39/42

page 36References.

Backus,D. (1993), Interpreting omovements n the tradebalance andthe terms of trade,' Journal ofInternationalEconomics34, p. 375-387.

Bean,C.R. (1986), TheTermsof Trade, LabourSupply and he CurrentAccount,' EconomicJournalSupplement96, p. 3846.

Breusch,T.S. & Pagan, A.R. (1980), 'The LagrangeMultiplier Testand its Applications o ModelSpecification n Econometrics,'Review of EconomicStudies 47, p. 239-253.

Bruno, M. & Sachs J. (1982), 'Energy and Resource Allocation:A Dynamic Model of the DutchDisease,'Review of EconomicStudies 49, p. 845-859.

Corden,W. M. (1984), Booming Sectorand Dutch DiseaseEconomics: Surveyand Consolidation,'Oxford Economic Papers 36, p. 359-380.

Corden, W. M. & Neary J.P. (1982), 'Booming Sector and Deindustrialisation n a Smal OpenEconomy,' EconomicJournal 92, p. 825-848.

DeGregorio, . (1992), EconomicGrowth nLatin America,'Journalof DevelopmentEconomics39,p. 59-84.

Etherington,D. & YainshetA. (1988), The Impact of IncomeTerms of Trade for Coffee on CapitalGoods hnports and Investment n Ethiopia,' EasternAfrica Economic Review4, p. 48-52.

Gavin,M. (1990), 'StructuralAdjustmentto a Terms of TradeDisturbance,' Journalof InternationalEconomics28, p. 217-243.

Harberger,A.C. (1950), 'Currency Depreciation, Income, and the Balance of Trade,' Journal ofPoliticalEconomy58, p. 47-60.

Hsiao, C. (1986), Analysis of Panel Data.Laursen,S.& MetzlerLA. (1950), FlexibleExchangeRates and he Theoryof Employment,'Review

of Economics and Statistics32, p. 281-299.Matsuyama,K (1988), 'Terms-of-Trade,Factor Intensitiesand the CurrentAccount in a Life-Cycle

Model,' Reviewof Economic Studies55,p. 247-262.Morley,S. (1992), On the Effectof Devaluation duringStabilisationProgramsin LDCs,' Reviewof

Economics & Statistics74, p. 21-27.

7/29/2019 107507322_20041117185522

40/42

page 37Neary,J. P. & S. Wijnbergen (1984), Can an Oil DiscoveryLead to a Rcccession?A Comment on

Eastwoodand Venables,' Economic Journal94, p. 390-395.Obstfeld,M. (1982), AggregateSpending nd the Terms of Trade:Is There a Laursen-MetzlerEffect,'

Quarterly Journal of Economics97, p. 251-270.Persson, T. & Svensson L.E.O. (1985), 'Current Account Dynamics and the Terms of Trade:

Harberger-Laursen-Metzlerwo GencrationsLater,' Journalof PoliticalEconomy93, p.43-65 .Salter, W.E.G. (1959), 'Internal and External Balance: the Role of Price and Expenditure Effects,'

Economic Record 35, p. 226-238.Schmidt-Hebbel,K & Serven L. (1993), 'Dynamic Responses o External Shocks in Classical and

Keynesian Economies,' CEPR ConferencePaper.Sen, P. & Turnovsky S.J. (1989), Deterioration of the Termsof Trade and Capital Accumulation:a

Re-Examination f theLaursen-Metzler ffect,' Journalof IntemationalEconomics26, p. 227-250.

Svensson, L.E.O. & Razin A. (1983), The Terms of Trade and the Current Account: the Laursen-Harberger-MetzlerEffect,' Journalof Political Economy 91, p. 97-125.

Warner,A. (1992), 'Did the Debt Crisis Cause the InvestmentCrisis?', Quarterly Journal ofEconomics 107, p. 1161-1186.

(forthcoming,1994), Did the Debt Crisis or the Oil Price Decline Cause Mexico's InvestmentCollapse?'.

Wijnbergen, . (1984), The "DutchDisease": A Disease AfterAll?', Economic ournal 94, p. 41-55.(1984b), Inflation,Employment, nd the Dutch Disease inOil-ExportingCountries: A Short-

run DisequilibriumAnalysis,' Quarterly Journal of Economics,p. 233-250.

7/29/2019 107507322_20041117185522

41/42

Policy Research Working Paper SeriesContactTitle Author Date for paper

WPS1384 Fiscal Decentralizationand David Sewell November1994 G. LangtonIntergovemmentalFinances n the Chdistine . Wallich 38392Republicof AlbaniaWPS1385 FiscalFederalismDimensionsof Tax Robin Boadway November1994 C. JonesReform n DevelopingCountries Sandra Robens 37754Anwar ShahWPS1386 EU Bananarama ll Brent Borrell December1994 G. llogan33732WPS1387 Fiscal Decentralizationand the Size Jaber Ehdaie December1994 C. Jonesof Govemment:An Extensionwith 37699Evidence rom Cross-CountryDataWPS1388 Does VoiceMatter? For Public Samuel Paul December1994 B. Moore

Accountability,Yes 35261WPS1389 Ruble Overhangand RubleShortage: Patrick Conway December1994 L. SukiWereThey the SameThing? 33974WPS1390 Regional Integrationand the Baltics: Pirita Sorsa December1994 J. NgaineWhichWay? 37947WPS1391 Where in theWorld Is Population Jeff Kling December1994 S. FalonGrowthBad? Lant Pritchett 38009WPS1392 Some Economic Consequences f Jean-PaulAzam December1994 C. Jonesthe Transition rom CivilWar o Peace David Bevan 37699PaulColrierStefan DerconJan GunningSanjay PradhanWPS1393 The Interface of Trade, Investment, J. Luis Guasch December1994 J. Troncosoand CompetitionPolicies:Issues and SarathRajapatirana 37826Challenges or Latin AmericaWPS1394 MacroeconomicReform and Growth LawrenceBouton December1994 R. Martinin Africa: Adjustment n Africa Christine Jones 39026Revisited Mi;uel KiguelWPS1395 A Typologyof Foreign Exchange JanineAron December1994 R. LuzAuction Markets n Sub-Saharan Ibrahim Eibadawi 39059AfricaWPS1396 ForeignExchange Auction Markets Janine Aron December1994 R. LuzIn Sub-SaharanAfrica: Dynamic IbrahimElbadawi 39059Models for Aucton Exchange Rates

7/29/2019 107507322_20041117185522

42/42

Policy Research Working Paper SeriesContactTitle Author Date for paper

WPS1397 Are PrivateCapital Flows o Uri Dadush December1994 J. QueenDevelopingCountries Sustainable? AshokDhareshwar 33740Ron JohannesWPS1398 The Cost of Air PollutionAbatement RaymondS. Haytman December 994 E. SchaperDavid Wheeler 33457ManjulaSinghWPS1399 How Important o India's Poor s the Martin Ravallion December1994 P. CookUrban-RuralCompositionof Growth? Gaurav Datt 33902WPS1400 Technicaland Marketing Support Brian Levy with December1994 D. EvansSystems or Successful Smalland Albert Berry, MotoshigeItoh, 38526Medium-SizeEnterprises n Four Unsu Kim,Jeffrey Nugent,Countries and ShujiroUrata

WPS1401 Colombia'sSmalland Medium-Size Abert Berry December1994 D. EvansExporters and Their SupportSystems Jose Escandon 38526WPS1402 Indonesia'sSmalland Medium-Size Albert Berry December1994 D. EvansExporters and Their Support Systems Brian Levy 38526WPS1403 Small and Medium-SizeEnterprise Motoshige Itoh December1994 D. EvansSupport Policies n Japan Shujiro Urata 38526WPS1 04 The Republicof Korea's Smalland LinsuKim December1994 D. EvansMedium-SizeEnterprisesand Their Jeffrey B. Nugent 38526Support SystemsWPS1405 Growthand Poverty n Rural India Martin Ravallion January 1995 WDRGaurav Datt 31393WPS1406 StructuralBreaksand Long-Run Javier Le6n January1995 R. LuzTrends n CommodityPrices RaimundoSoto 31320WPS1407 Pakistan'sAgriculture Sector RashidFaruqee January 1995 F. WillieIs 3 to 4 PercentAnnual Growth 82262Suslainable?WPS1408 MacroeconomicManagementand Jun Ma January 1995 C. JonesIntergovemmentalRelations n 37754ChinaWPS1409 RestructuringUganda'sDebt Kapil Kapoor January 1995 E.SpanoThe CommercialDebt Buy-Back 35538OperationWPS1410 MacroeconomicEffectsof Terms- Nikola Spatafora January 1995 J. Queenof-TradeShocks:The Case of Oil- AndrewWaner 33740ExportingCountries