DIRECTIONS IN DEVELOPMENT
>~~~D c Zq _ .
Ma raging ico a t-Qdi- V-Boo
">and BustsPANOS VARANGIS
TAKAMASA AKIYAMA
DONALD MITCHELL
rz. . . ' ' t-V ~ _ R E!_
- 4
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Contents
I'oreword v
SLtmmary vii
Wlhy boomils happen I
Wlhy this boomii won't continiue 3
Managing booms an(d busts 8
Learning from experience I()
Who manages the boom? 17
How to manage booms 19
Referenices 21
DIRECTIONS IN DEVELOPMENT
Managing Commodity Booms-and Busts
Panos Varangis, Takamasa Akiyama, and Donald Mitchell
The World BankWashington, D.C.
1995 'I'he Interinational aBank for ReconstrLuctionand Development/ TI IF WORlID BANK1818 11 Street, NWWasiiingtoin, DC 20433
All rights reservedManufactured in the Ulnited States of AmericaFirst printing Decemliber- 1995
't'he findings, interpretations, alid con01CLIsiolIs expr-essed il this studv areentirelv those of the authors and should not be attributed in am1V man-ner to the World Banlk, to its affiliated orgniiizationls, or to mcinbers olfits Bioard of lExecutive Director-s or the counltries they represent.
Lilbr-ar1i (1oi ( ,dfnIOgrct<-iIt-pnbIit(,i(m s)au
Varangis, IPanavotis N.Managing commodity booms-and busts / lanos Varangis,
Takamasa Akivama, IDonald Mitchell.p. cmT . (Directions in development)
In-Cludes bibliographical referenices (p. 21)ISBN 0-8213-3489-11 C'ommoditv futures-Developing counitries. 2. Primilarv
cornmmodities-Prices-Developing countries-Forecasting.3. Export malrketing-Government policy-Developing countries.1. Akiyama, 'I' ('Takamasa), 1944- . II. Mitchell, D[onald, 1947-Il. 'Iitle. V. Series: Directions in development (Washington, D.C.)I 1G,6051 .D44V37 19)95332.63'28-(dc2tt 95-42336
CI P
The authors are grateful to all those who helped in producing this booklet,particularly M. Ahmed, 1. Cuddington, A. Kraay, S. Lateef, J. McIntire,Y. Mundlak, J. Salop, A. Ray, and M. Schiff for their useful comments. Theauthors bear sole responsibility for the views expressed. The editors wereMeta de Coquereaumnont, Paul Holtz, Vince McCullough, and Bruce Ross-Larson, all with American Writing Corporation.
Foreword
In developinig coulitries dependenit on a few commodities for thtirexports, boominig commodiity markets can be beneficial or detrimeintaldepenidinlg how they maniage the windfalls. Iwo key priobleiiis need tobe addressed: the fluctuatiolIs in incomile, and tranisitory bonanizas inforeigIn exchange earnings. '[Ihis sccond(l effi'ect. in partiCu-.lr, is likely tocause tht ex(han1ge rate to appreciate ahove its long-term sustainablelevel
'[he lessons frome CXpric-ilCC are clear: ounlltries that have managed toinIVest the gailns from the hoom come outI oft it stronlger and with highierrates of growth. ( ouIntries thalt, throtgigh misguided policies, Con1Slime'the windfall gains aid allow prodLCtioll to shif tfolml exports to (dolimeS-tic markets have n otIhinlg to slhow for it at the enidc of the boom and(i find(it diffictilt to\ tinlto the IiSali-11gmIlent in rIlative prices produced by theboonv
'his bookiet focuscs oll the management of comimilloditv price volatil-ity t'r-omil the perspective of developinig countries. It examines and conl-trasts goveriin menit policies anid institutional aldid mi-ketinig strLlctul-sthat have p1oveII to he effeftive ill mnanaginig com modity booms-andliuIsts.
Michael BtrulioSenlliol Vice P'resident, Development Fcollomilics
and1 (i Cief [Lcolloiiist
Summary
icO;iouing comml1llodity iimarkets can hrinog econoimic good o0 ill to devel-Oping couLntries that are exportinig these commilodities, dependinlg onhOW theseS COUL1tries halade thte windfalls. Su(ch booms occur roughly
once everly t'll years. 'I le latest began in 1)992 and, while most comi-modity prices have probabl)y peaked, developiing coulitries still have anlopportulnitV (alheit brief) to imilprove their fiscal positions while prIicesreailnill Iigi.
Ihe ecollomilies of detvelopinig counltries are heavily dependent oncommiliodities. Primnllay COMMOditieS, inClUdilig elnergy. aCCOUnt for, onlaverage, niealy Ic- -haIf of expi ort revenies frFoim developingc, countries.Ainy booml in comIlimodity pirices Imleanis a bo mn in governi menit revenues.In the past, policvymakcrs have ofitenl mismanaged the' boom. Whelnpr-ices f'all, these c(oinoimies are lef-t in w*OIse shape t'than before.
Whell revenLues Itrii coni C im oditv exports increase, goverinenlits aretemllpted to mlake long-term spenidilig coilnilitlinenlts based on whiat turniout to be short-term pIrice rises. Witness tColonlibia, Kenyva, and l.nizaniaaflter coffee prices boomed in the late 1 970)s. In the post-oil booImlNigeria ol the I 97'0s and early l')X()s, theI ShortfallI of rwevnlues was sosevere that eXteFrna,l borrowing had riisein dramatically eveil befole oilpriCeS be'gall to fall. It was ImLuchl the samec stor-y in Vee'ZUela's andMexico's hanidlinig of thetir oil boomis.
Recent research sIvOWs that fluctuLating (C.DP and inlvestmilenits are detri-mental to long-terin econioic glOwth. I len( e liscal policV should beCOuntrClvCCiCcI. (COVernl1llelltS experiencing boomns, hlowever-, tenid to
adopt a procyclical fiscal policy becaLluse of the windfall tax rIevlue.1oFreover, foreign exchiange eariii[ngs alImost always soar dIl-ilig times
of highler commillodity prices, which canl lead to an excessive appreciationof the real exchange rate. 'libis couldl make somile tradable sectors lesscolimpetitive in global markets and, ultilmlately, lead to a decline indomilestic prodUCtioln in other, nonhoominig sectors. '['lis is solimetimliesknown as Dutch disease, after tht experieinice of thie Netherlanids in thIl')(.Os followinig the discovery (and expor-t hoom)) of Northi Sea gas.
Anlother problemn associated with commlilioditv boomis is a slowdow l iindiversification. Since proIucers often hIve few inivestimenilt alternat yes,Windfall r-eVCnluICS are inlvested ill the bO0minl1g ( onllmlloditV SlibseCtols-
%6
viii ,\ANA.JN(. (I()%1MNI)11) BiOONIS-ANI) 1311SIS
even wheln producers are aware that the boom is tetmiporary. 'T'lhis calllead to overcapacity and much lower prices later anid can have lolng-termiieffects on a cotIntr,vs diversification elforts. Liberalizing and developingfinianlcial markets canl provide altei-rnative inivestiments. l I nderldevclopeddomestic financial and capital markets, however, may niot be maucIh of aprioblem it prodLucers canl invest freely abroad.
Trade policics cani also par-tly dleterminieii the effects of a comlmllloditvboom. If they preveint revenllies I-roin being spent oni imports, traderestrictions many lead to fuLr-thier- apre( iation of- the exchange rate.
Learning froimi experienice
In the past, governiinetits introdLuced varioLus policies to mianage booimsand avoid thle onset of DLitch disease- I[lhey haive followed prudent fiscalpolicies, lifted impor-t and capital restr-ictionis, tightened monietary poli-cies, accumILulated foreigni reseives, and rediuced foreigin debt. lut whatwere the specific measures, and how etffective were thev?
Ixp()I (IxCS. InI tile latest coffee hoom, ( ameroon, (,cte dlvoire,Honduras, lIgaida, \iet Nain, anid others reinltiodLuced export taxes. SuLChtaxes, so the argument goes, can alleviate both the spending and theresource movement effects of tile hoom and protect noinihbootminig trad-able sectois. ' lax teveILIes should be invested in high-yielding publicinvestment programs that inicrease long-term growth. Maintaining highstaU dards of pi ject eva I-nat ion is also illpo-talit.
Inl pri`actiC, hll )eCVei-, talX rates lre uaLIS1l,1y set too llighl an1(d pLiblic inivest-ment p-rojects ale oftell selected on n1onetCOnIOmliC grOulids. Ib'is has dla-aged the bo omiiiig sectlor with little cominpensation elsewhere in the econi-oniy. I gligh export taxes also enlcloul-age tax evasioni and smugglilg.
Commoility bo,mn lyoids. 'Ihese are bo( nds denominated in lforeign cur-renicv hel(d bv local expor-ters of cominmodities. 'Ihe bonds canl be conlsid-ered as forced sa\vings in l'oreign curr-ency. I heir purpose is to protect theeconollmi of the coinmmodity-exportinig c ouItitV hroml the effec(-ts of a sLid-deni inflow of loieigin excChainge. For example, dueL to a rise in conlinod-itv pr-ices in Colomnbia the cenltral bank issued "coffee bonds" toColombiahii coffee producers in exchInIge for fol-eigil (curi-ency eal-rle(dfrom exports. ( offee exporters were reCLuirled to hold( a certaill percent-age of their dollar-dlenominated revelnues ini these bonds.
IHeidgin-' inst(uinelis. Covern ments and t[ ) private sector- can makefLutlure revelLnus mnore pre(lictable by tisiing options, futLl-es, and swaps.'[he NMexi(all goVernlmllelit used sLIcI hIedgilng tools to eCIsure a gLaran-teed price for its oil (and predictable revnTIues) after the Gulf war.'I raders and producers, too, could benetit froim hedgiing. Some develop-ing co( ntries would haive tiouble in using hedging i nStlu iLeltS, hoWeV-
er. Also, hedging c-ani he costly, requireS g(ood cr-edit ratinlgs, and is moleeffective wheni used to cover coininCodi ty prices lor a relatively short peri-od( of tililC-a1boLIt a yearI1 for m1oSt (olinn1o(litiCs.
MiNAR15 ix
Rev'n li stbilti liofltm flintlls. Aniother- wav to manage commiiilodity priceshocks is to create a stabilization f'Land to stabilize gocvcrnmilent expendli-tuL-es (not to suppro0t or stabilizc comillodlitv pirices). By savinig resouIrcesill boom yealrS, governl1me'nts can redu-e the nteed to CUt Spellding in
lealer times. A well-dcsigned fuintd canl smloothi expenlditulre adjustmentsupward andl downward. I lowever, fuindis often] runL ilnto difficulties
becauIse of' the u rlpedictable nature of cornoniodity prices or in isilmall-ageCImenlt of1 tunid IreSOLur'CeS.
Init'rat7 until( ( 0oniinCdy ,I(7t'ntC1tt. [iUring the late 1 97)s and the I 98tisanyliv c oin modlity-pCrodtucilng coun1tries altemClpted to stabilize comnmiiodi-
ty prices (colicc, (ocoa, aind rubber, For linstanIce) thlrOughl ilntelrnatiOnlal
agreelnets. ( )nIv thle uibber agreement is still In IOrce-though barely-and its effec(iveiless is bCeilg (LuStiolnCe byN soIe Sitgn,atories.
I lowx to mlaniage a boom]
Commodity priices are especially hard to prIediict. Crops (and(i so prices)can be affected hb weathier, insectS, alld (iSCaSe. And world prices can beinfltuenicedl by sp-e)culators and political e venits in exporting andt imipl ortinig
countIries. SLIc unllle(trtailnties miiake anvy ki(i of planning difficult. FVenI so,experience has shown that cotintries that adopted prudielt fiscal andillionetiA' pOificiCs, ccumllateldlCLI reseIves, redu(ed fo`reign clebt, and liber-alizedl capital contitols alnld import restrictions saw ilong-term bcneficialresiIIts f'romctin (ommoditv' booms. ()ther, coiminpleillenitarv policies have alsobeen SuIggeteSdL. TCeSC policiCs COUld provide addi tional henefits, depend-
ilig on tic ( olitions of the ccoloilrliCs c ollCerled 111an the nlatilre of the
booil l olih ies mliloloyed to manage coniiniiotditv windfall revenlutes are:
l)(1tn ' (1 tii't' IIIt/ vI tlI'ltlitII. rI. ( ;vOernivi IcIItS should0 ILot COIoiiI it to 10log-
terili spendin(ig levels that they Canno110t SLut ni VsliLn thl th bOOMl fizzles.AdOPI 1"'s'/Uzt (iu', mum)''iV *1tFil 11/tl p/,1ht 1c. (t :omlltod ity-depcitdent
Ceon1ollaies are milore vutlnerable ti (exte r Iil shocks, making gnioodl macro-
ecollomlic policies especiall\ (c-itical1. t CSItiOes shlId follow pruCenItmonetary aid fis s1l policie, alnd theI\ ko(uld Wevell CoilSidLer Redul(cingexterna,.l dbLt with the proceedS 1from11 the boom. ll 11m.1a1ny CaSeS a ri0gidbuldCge't1a ' I)IrCcsS Cl ri ig a hOi o111 C Ia reSult ill illefflCiCilcet uIse of tilhWinlltlil revI t LIn (as w 1he the rvI\ t ICC a IcS to 1o 0 11i i 11niStistr )
( oven-CHiL ts sI 11i lld coniside-r init mil'vimig thet bmdgetairy process doLl-ilngbooms. Also, govenlmeilts experiilcilig (oinlodity boom01s tecln toSpenILI tht viilldflAll tax revenutc fromii the boon I ll is procyclical fiscal
policy appioachl makes a (timLlilti-vyS ()l) flu(ctuatel 111m1or-anld thiutt'
shouIId he avoided.R'ducti oilfrol-,. (Capital-acCoUilt conttrols ( restrictions oni itl\esti :
abroad) ani]d i IpOrt conItrols sIoutL1d be r-edLill(Lc. Peolicyinakeis slikt I
con0sider takilg advantage 0if the faVOIrabli tCrde al_aI ce sit iatioll to Ise
ili port restrictiolls. 'I 'li is will help avoid a real ex( hl nge rate appreci' iOIl
a 11(1 v ill imi provcecitnltlin itrf(iciicy. I'li1en, whleni the ltho tli ends, thle
x NIANV4\(,IN(; (:)NINMoI)II[ I,()()\ D\NI) IrS i1
econoimlv will he in better condition than it was before the boom.Diversifi' and adjust. It is less paiiifitl to begin sucIh programs when
prices are higil, and booming sectors will be better able to cope withlower commodity prices wheln the boom subsides.
A r-ellentueit sllii'z.atiafn jund Illav hbe an optioll Under somIIe CilrCLIum1-stances. It can smooth adjustments in spending wheln the hooml is Over.Funds can be USeful whell dolmlestic finalIcial markets ai-e not well dlevcl-oped arid there are significanit adjLIustIlInt costs to booms and busts.Hlowever, tuIln(s ofteli eIcouLIter financial diffictilties anid are sLbIjct toim rsmianagemen it.
Hedge. [he use of market-based hedginig instrunments (fultules, optiolls,swaps) mikes conmmodity-related revenLues mior-e predictable, increasesthe probability that anticipated reventues will le realized, and improvesthe abilityi of policyrnakers to plan. I Iedgirllg instrum-1ent1s Call be usedlindependently Or r1 tcomCnpleImlent a stahili/ation fUnld in order to ironOUt flows ilto andi tt'oIml the fuI(l.
Avoaidl 11c'a,ri c(xort talx-es. '[he'liet n11ay eInCOUrage tax evasion anldl sImug
gling. Nol-eover, governlmlellnts tend(l to speiid increniental tax reveCLresswiftly arind poorilv. low to nioderate expor-t taxes (acrinig as a windfallprofit tiax), however, c(-ulnd have benelicial ef'fects in reducinlg ove irivest-melit ilr the boonlii ng sectoI antd providinrig iniceritives for- diversification.
l3ecauise the track record of(govern nirients in mnanagiig booms and bustsis farfr- fi-onln iilllessivo, the private sec(to shoukl be allowed to mainage alarger share ol windfall prolits. I lowever, the mantagement of windfallprofits by IlIrOUlUCCS 1ll.tV still be frrstrHatMd hV dlistortionls suichi ias capitalaccount o nitrols, trade restrictionls, arid macrnoeconominic instabilitv. lofacilitate privale ralnagillrelit, govelrlill(lits should i-erinove these distor-tiions by pimtsuinrg trade and financial liheralization.
Why booms happen
Booming commodiity prices can help or hurt developing-countryeconomies that export these commodities, dependinig on how govern-mncits handle (or mishanidle) the booml anid, more particularly, howthey spend the windfall export revenues.
Althougih most comimilodity price fluctuations are subject to long-run cycles, there are also short-term variations (booms or busts)whiose effects are in essenice quite similar. Comimi(odity booms happenevery ten years or so. The longest and stronlgest booms were the twoin the ] 970s, when prices rose bv 240 percenit in nominiial terms; themost recent boomn began in 1 992. Andt altiloLugIl there are some sim-ilarities between the 1970s and thie latest price increases, there arealso some imlportant differences (box 1 ).
Thle recenlt surge in prices appears to be based on supplv shocks, notonl incr-eased demaid. This suggests that prices will fall as productionreturins to norimial or highier levels. AlthougIl many commoditv priceshave probably peaked, developing countries still have an opportunity(albeit brief) to improve their fiscal positions before prices fall.
Most exporits of developing couLntries are in primary commodities.Includinig energy, these accoulited for 76 percent of Africa's merchlian-dise exports in 1992 and 47 percenit of merchandise exports for all low-and middle-inlcomile couLntries. For severely indebted coulitr-ies, the fig-tire was 61 percent. Iuels, minerals, and metals make Lip aboLut 29 per-cent of developing-counltlr export earninigs, with agricultural products(includinig timber) accouniting for anotiler 18 perceit.
Thus developilIg couLntries rely on comimii(odities for much of theirexport eariniigs. The latest boom has favored these counitries becausethe prices of the comimillodities they export (such as coffee, cottoin, andcopper) have genlerally risen more thani the prices of those that theyimport (grains, petroleum, and so on). I ligher cottoni prices benefitedcouLntries in Sub-Saharan Afirica and( Central Asia; Latin Americanlexport revenues rose on the back of booming copper prices and highiercoffee prices, whicih also benefited Sub-Saharan Africa.
Llncertainities associated with commodity booms mlake any kinid ofplanniniig difficult. Bliulper years are oftenl followed bv busts that typical-lv leave real prices lower thanl before. When revenLues fromli comilmodity
2 , I \,i \(.IN(.(:( )PNRMOI 1) Boo)M\lS-\N)D lithI,
exports increase, govern ments are tern pted to make long-tern spendidigcoimlaitmeilnits based on whiat turnl out to he short-term price rises (as
happcllene in the 1()70s). Moreover, the increases in For-eigni exchanigeearninigs ofteni associated with higher coin iodity prices call leadi toexcessive and proloniged-appreciation of tlhe real exchange rate. I'hiscoLlidl mlake somie tradable sector-s less com petitive anid, Litimilately, leadl
to a deciliiie in domestic prodCticnio iii othelr- sectors, notably nianiufac-
tUril1g. 0ommoil)ditV hoolns also hold ha k dliversificationi into otlher-export collilliodities.
Box 1. Boom, boom
The recent commodity boom differs from that of the 1 970s in several ways.In 1973 the prices of two of the most sensitive commodity groups (grainsand energy) increased sharply. Although grain prices have risen this year aswell, their 25 percent rise is far below the doubling of prices that took placein the 1970s. This might be why the current boom has not attracted muchattention. Moreover, the Soviet Union, which imported large quantities ofgrain in the early 1970s, no longer exists, and the import demand of thenewly independent states is severely curtailed by foreign exchange con-straints and weak economies.
Fertilizer prices have risen recently because of production capacity con-straints and sharp increases in demand, but by less than in the 1970s. Thismeans that the crop yield effects of higher fertilizer prices are less today thanduring the 1970s. In the unlikely event that petroleum prices move sharplyupward, the increase in fertilizer prices should be reversed as productionresponds to higher prices.
Another difference between the two booms is the role of commodityfunds. The emergence of large commodity funds in the 1980s has perhapscaused some changes in commodity market fundanmentals. The emergence offunds has increased competition among those willing to hold commoditystocks and has increased the arbitrage possibilities in the market (pricesrespond faster to events). At times, funds may have accentuated commodityprice trends. Some analysts have found an association between fund activitiesand the level of prices during 1994 for certain commodities, such as cocoaand metals.
Although the price rises for some commodities have been large, theyscarcely compensate for the declines of the past decades. Aluminum prices,for instance, have yet to approach the highs of the mid-1980s. Coffee, whichhas more than doubled in price in the past year, is still at less than 30 percentof the peaks reached in the 1 970s. Cotton, cocoa, and copper prices are aboutone-third of the highs of the 1960s and 1970s in real terms. And while petro-leum prices have increased recently, they are still below their 1992 and 1993levels.
Why this boom won't continue
A commoiditv booml is a sharp prict r ise in a number of comnioditiesover a relativelv shiort period of time (t) ull to about two to tlhree years).'I[he increases sinice late 192 91have certainly been sharp and broadlyhased. Ilhe lWorld 13ank's indlex of -33 noniluel pri mary commodity pricesrose 42 percenit, fromn a low in Octoher 1'9)2 to a hiigih in March 1995(figule I and table I). Food pirices increased 21 perce nt, aind the indexof- agricultural raw materials was 54 percenit hiigher. Metal and minieralprices did niot increase significantly until late 19'93, whell they bounlcedupward, increasing 41 percenlt thiougil January 1995.
Figure 1. Weighted index of primary commodity prices forlow- and middle-income economies
Petroleum (spot) Nonfuel commoditiesIndex: current US dollars (1990=100) Indiex: cur,ent US dollars (1990=100)
I -lo I
Total food Metals and mineralsIndex: current US dollars (1990=100) Index: cLirrent US dollars (1990=1001: I_
1S-', &4.1 ^ I:S' 1
'K- 5-.: 2, 5 I'' t 1" 1 -r, L1,- 1-., P, ,-. ' S I I"VJ1 I I - I r11'I
,Jotr o-fl tn,-i dllO meletas aiid lii rewais are ficludeo | the w ,lftjei -rpnicinda es 1ide, Fr'n - sTieotlwefod bM 4 el tr UrfLel ,lrncdiles ndp- ainUptals anrl nlerals by 28 14e inde CurrII1 Sdo tl l'a (1990 a 1,d IndT,alye, x: l-t lnrteren, l 1 U dollars'` 1990 v100cr d B)k
4 MANAGING (C(MMODIIY BOOMS-AND Bt]S-IS
Why the recent surge in prices? Declining commodity prices through-out the 1980s led to reduced investment in production capacity, which
Table 1. Weighted index of primary commodity prices for low-and middle-income economies in current dollars
1990= 100
113,, ,,1,,,.N,/ue tl (r-l R(i h , (,,) 1,il', Ml,ls
,dilie X,1 1I/.l (,?i, 10t1,1il Omw 1 . 0s1h.e,.,<.. Bevolgl ,,,1,,tjIil lowtil;,
I'l',lilmll (11(6.90) 1(19.1) (29.4) 1t.9) (1luI) (2 2.4) (M 9) (22.8) (1) i) (28. 1) (2.7)
Annual1980 161 1 126.3 136.9 139.2 1 34.3 148 6 134.3 185.1 104 3 79)0 95A1 128.1
1981 155.1 108.6 116 3 122.9 143.3 141.7 96.0 148.8 89.9 88.4 82.3 122. 31982 142.7 95.9 104.0 96.6 105.7 117 .4 74 5 149.5 79 9 66.1 75.2 104 .1983 129 6 103 3 112.3 105.4 109.6 135 6 7. 2 157.2 87 9 63.7 81.9 96.11984 124.6 105.1 118.0 116 9 104.1 157 7 66.7 179 6 67.0 68.1 74.1 97.71985 118.8 91.7 100 5 86 3 89.2 113.0 62.9 165. 3 70.8 (0 1 70.2 81)01986 62.7 92.6 103.8 77.1 7(. (6 87.7 68 6 1 93 7 70.5 6 3.9 ) 65.4 69.3 1987 79.39 2.9 98.8 84.5 77 5 1(01.2 74 7 1 35.6 80.2 80.1 76 4 94.41988 64.3 111 3 110.1 107.3 102.2 133.7 88 5 141 2 9116 80.4 114 6 10871989 78.0 107.3 1i)8.0 107.9 112 0 119.1 96.5 1150 9 6.9 9.12 111.3 10(6.21990 100.0 100.1.0 1oo (3 In1.0 10o0 1130.0 10(1.1 1(1(1() 1011.(1 1111(1 1(1(1(1 10(1(1
1991 84.7 95.5 9 7.9 19. 2 101 7 1114.3 93 1 ) 3.6 9.1 1)4.2 869 1(02.41942 8:.31 12. 1 94.4 11)0.0 101 7 111.7 89 5 79. 4 )8.3 114.5 86 1 95S 81993 73.6 91.6 99.1 98.6 93 6 HI 5 90.7 84.9 110 3 152.4 74.0 81.71994 69)4 111 9 1213.7 106.8 102.1 126( ) 93.8 1504 1258 1 56.6 84.6 93.4
Quarterly2Q94 707 105.4 116.5 102.8 96.6 122.7 89.9 126.2 127.0 163.4 79.3 )3.1
3Q94 74.4 122 2 137 ) 105.8 95.4 126.1 15.1 201.5 132 2 163 ) 86.6 93.640Q4 72 1 122.5 133.3 110 8 101.9 135 9 )5.3 177.8 129.5 149.1 98.5i 96.3
IQ95 751 1 26.3 1364 1134 104.9 135.1 101.4 1692 141.9 142 1 103.6 101.72Q95 79.3 1 24.2 134.9 112 6 110. 5 1 31.0 98.8 163.7 14 .2 143 1 11(1 .2 10 2. 63Q95 71.9 120.5 128.0 1 1 9.5 128.4 136.1 100 9 145 7 125.9 1 18.5 103.7 (112.6
Monthly1994 Ian 61.8 95.8 105.8 109.2 121.8 122.7 91.0 92 5 111.1 147.9 71 9 899)1994 Feb 60.2 98.1 107.8 109.3 117.8 116.2 97 5 95 7 114.8 146.1 74.9 90.21994 Mar 59.5 98.6 106.3 105.4 113.8 117.0 96.9 100.6 117.8 148.8 754 91 I1994 Apr 65.9 99.5 109.9 10:3.7 101 8 118.0 93 I 102.9 123.2 157.8 74.6 92.31994 May 71.2 105.9 117.2 102.7 95.3 123.3 900O 128.6 127. 3 164.2 79 5 93 61994 Jun 75.0 110.7 122.4 101.8 92.7 126 8 86 6 147.1 130 5 168.4 83.9 93.61994 Iti) 78.1 121.9 137.5 1026 90.4 1206 94.6 20.3.3 133.7 167 3 86.5 93.61994 Aug 74 3 120 2 135.7 105.8 95.9 126 8 94.1 192.2 132.5 167.7 84.7 9 .61994 Sep 70.8 124.6 140.5 109.2 99.9 130.8 96.7 209.0 130.2 1627 88.7 93. 6
1994 Oc( 72.0 121.) 135.1 1116.9 101.8 1285 92.0 192.9 128.7 152.9 92.2 95.51994 Nov 74.7 122 5 132.6 110.5 101.2 138 8 92.5 177. 5 127 9 149 9 100.3 96.7
1994 Dec 69 7 123 1 132.3 115.2 102.8 1404 101.4 163(0 131 8 145 2 1029 96.71995 la(n 73 6 126.1 1 34. 3 III 7 104.6 134.7 96.9 166 5 139.7 147 8 108.5 10(. 11995 Feb 75.8 126.0 136.6 115. 1 104.7 135.2 104.5 166.6 142.1 141 4 102 2 102.61995 Mar 75.8 126.6 138.4 113.4 105.5 135.4 9'9.8 174 4 143.9 117.2 10(1. 1 102.61995 Apr 81.5 126.3 137.5 111.2 104.4 131.0 98.7 170.5 147.2 141.4 1011 0 102 61995 May 80.5 124.6 136.4 111.5 108.8 129.6 98 2 167 1 145.7 143.8 97.7 102.61995 Jun 75.9 121.9 130.8 115.2 1182 132.6 99.3 153.5 134.0 144.5 101 ) 102.61995 Jul 75.9 121.6 129.2 121.7 127.8 138.5 104.5 145.0 127.0 141.8 10.0 102 .61995 ALug 72.0 121.3 128.6 118.2 125.4 134.2 101.1 152.9 124.1 138.3 lOS I 102 61995 Sep 73.5 118.3 126.1 118.6 131 8 135.6 9 72 139.2 126 1 135.3 100.6 102 6
Note: Weighted by average 1987-89 export values for losw- and middle-income economies.Source: Coiiimodily Policy and Anialysis L1nit, InternaIioIial Econiomilics Department, World Bank.
WHIY IIIIS 0() ()\1 WON (I I .N I III S
led to supplV deficits in tihe carly 1990s. Ihis was followed by poor har-vests, and hbginning in late 1992 prices began to climb. Despite therecenl downtuLn, prices COLt1i mOeIove higher in thec short ruLn. Ihe longer-tCIeIr PiCtUIe. howeveit, is less thIan rosY.
'Ihe World 13ank's index of prinlary conmmliodity pirices is expected toilcI-rase by 8.6 p(ercelt ill nlom0linal tCrmIIS in 1995 an d then decline in1996 and 19)97. In real teriis, it is projected to declinie 8 percenit by 2000froimi the March 199)5 peak anid remiiaini near- that level for the next fiveyears. ( Cotton prices are expected to rise another 1 9 percelit in 1995; thisfollowvs an increase of 18 percenlt in 1994. (rain pirices are expected to;ise in 1995. with 11ma1ize, rice, an1(d wiheal all higher.
Apgic-niitio(l ('out ntodiii Iit b'n oom in agricLi ItUral prices began withriCe and 111 unusuial ly po01r harVest in lapall in the SUmnI1er of 1')93.Japancse rice imports rose frorm zero inl 1992 to 2.2 iiiilion toIns in
1993. World prices almost doubled but fell just as sharply whenll Japan'sproduCtion recovered in 1 994. Also in the sum111 mer of 199 3, the t liitedStates maiZCe CIrO) was slashed by a thiird because ot floodi;lg in theMfidwest. Maize pr-ices soared bLUt lost all that ground bv November1994 as lUS productioLn recovereFd.
It was the same storv of poor harvests for other comimlodities. Insectdlamiage redcticed the world s cotton crop, while (;ubha's sugar productionconltillued to declinie. [our maijor cottonl producers-China, India,Pakistan, and the HInited States-reported mucI lower crops in 1993.World productioln h as recovered substantiially since, buLt disease and pestproblems in China anid South Asia are delayilng a full global recovery.
In 1994 SUppIY disruptionis also extenided to coffee. Brazil lost an esti-m1atecd 40 plccrcet of its crop because Of frost and drought. At the sametiimle, world coffee stocks were at their- lowest level since 1979. Pricessoared. Cocoa prices, which had fallen by 70 percen1t in real terimis from1984 to 19992, have increasedc, too-by 60 percIenIt from 1iid- 1993. Cocoaprice increases were not due to weather disruiptions but to adjustmenits indemanid and1 supIply folloWinlg a period ofverv low prices. Still, cocoa pricesare very low by historical standards.
AM1etals e(1111 nuiilc('rals. An increase in global demiiand and a reduction inworld stocks have pLushed metal and milleral pirices higher. Voluntary pro-ductionl cuthacks aned restructuring in miining, processing, and fabricationhave also tightened supplies and increased prices futhier for some metals.I I ighel metal prices have bee n partly credited to speculative buyilng by com-modity utnlds. If so, the boomil in prices will be short-lived and may alreadybe over for iallny metals.
Crude oil. Aftel shortages durinlg the (;ulf wai', oil supplies picked upin 1993 and prices fell significantly, before rebouLn1lilig in the secondqIuarter of 1994. World oil demilanid will probably grow modestly for therest of the 1 990s, and crude oil prices al-e ulilikely to rise much, espe-cially if Iraqli crude ree nters world markets.
6 NI\NV,IN( ( MiN1 M(WIIl IBOMNiS ANIl) [itlNIS
TIhe revenue windfall
For maniv developinig couLltries, priniary comimiodities still accounlt forthe lion's share of export re-venues. In 1 3-94 the average price of pri-mrary comiliodity exports l-ose 1 9 pcccrcent. he effects on1 the revenLues ofindividual CoulItries was otCten mucIL greater hecause of the conrcentra-tion in a few boominiig coininodities. ( oriplete export revenLuC figuresare not vet available, buit the World Bantk has estimiaite(d revenues foreight agriCLiltural co(mmo1 1dities-cocoa, coCon1ut oil, coffee, cottoll,groundnlut oil, pal in oil, rubber, anid sovbein oil-anid three mietals-alulmlillullm, COppCe, anid nickel-for all low and mli(ddle-ilicomiie coIll-tries (table 2). Agri-icultul-al exports accouiiited for 70) percernt of the totalincrease in export revenLues, alld coffee for hilf the increase in agricultur-eexports.
'Ihe gains fromil highier- export prices were widespread (table 3). LatinAmerica and(I the ( aribbeanl notchied up the largest gainis, followed by lastAsia and(I tlhe Pacific. Latin America and the Caribbean benefited molstfrom tht heoooim in coftfee prices, althoughli other comimloldities, includlingmetals, also contribUted. For last Asia and the Pacific, the increase wasdue m<ostIV to hiiglher- prices for rubber, coffee, palini oil, antd other veg-etable oils.
Which countr-ies were the big wininers? In dollars, Malaysia registeredthe biggest gain ($1 .76 billion). Paliii oil led the way (S1.36 billion),followed bV rubber ($340 miilliol). Ihe other leaders were maiily coffeeexporters. Brazil benefited the most, despite its coffee prodtuction
Table 2. Export revenues from major primary commodities forlow- and middle-income countries(billions of dollais)
Comimiodit)i 1 992 1 99') 1 994 1 9() 3-94
Agricultural comnminodities 8.0 18.5 28. 2 9.Cocoa 2.2 2.2 2,7 0 5
CoCon1i Oil (0.8 0. 0.7 0).2(loffee 5.2 .8 Mt).6 4.9
Cotton 2 .6 2.4 2 8 (.5
GrouMIdniLt oil (). I O. 0.2 0. I
Palilm oil 2.8 3.1 4.9 1.8
Rubber 3. 5 4.0 1.3Soybeall oil (.9 1.1 1.5 n.5
Metals andl miericlals 14.4 1 V.8 17.8 4.0AluminuLm1 4.9 5.2 7.2 2.t(Copper 8.1 7.2 8 1.4
Nickel 1 .4 1.4 1.9 0).6
rS)im tp I stilli. tes 'At thi ( miMllli(V I oli(y .111d A1l.1is t inil. Intemitionakl .olono ic-s
Dt-pallrtme m1 Wodd 1liank.1
\V1I I 11 1 11 It 15 \\'()N I 1 \I IN \ I II 1
problems. Colobilaib, too. Inidon1esia Was also hlighi upl amon11g tIle will-ner'-S, with la.rg gains in export revellues fromil rIhu ($400 million),palm oil (5300 imiillionl), and coffce ($300 iillion). At the top of' thleexport revenue gainiers ill Suh-Sahaian AFrica was (C ote d'Ivoire, thaniksmainly to co(oa ($250 miiillioni) and coltec ( $200( millionl). Other CouIn-tries wNSithl larl-ge galillS ill export revenUeS inlcltide Rwanda (3 7 percent),
Chad (28 pci cern) and 'lhnzania (1 7prenmt). As a shari of( DI, theinlCreCaSe ill expi 0t IreVeIlLIu Was IlloI- than 30 pecent or lKt -liopia aindalmost 6 ertiitlft iol (ote di'Ivoire (table 1). Il laddition to SLIb-SaharailncoLnIlt ies sIvei l ct ti tv-dependent lif nl \IIIerica ll COLnitries bene-fitedl ionli til loomi, imaiilvkw tle toffee CxI)tOHL s.
Table 3. Export revenues from booming commodities, by region(billiolns oi tdollals)
Reiou 17992 1 ' 7994
ttast Asia andt ilie Pacific '1) 0 i 1 3.4I'lirope and ( entral /\sia 3 1 4. 3 (.4
Latin America aid the tCaribbeain t 1t 2.2 1 7.8Middle Fast a)id North Africa 0.4 0l 0.6SouIth Asia 1 .0 0 0 7Suib-Saharaia-^ti Africa 5.7 5.4 7. 2.to ctw I stimjss bv ii (I omrn iohj v totics anid Anilvsis I Iii, Initei nattitl tLioillmics
I)Cpartmlentl, Wm'd,% ,1(1 1.,,
Table 4. Big winners, ranked by primary commodity exportrevenues(billiolis of d(llars)
11io-eist' as(omantSt11' 7992) 799'S 1"'994 '"-)4'b of 93 GDP
Malavsia 3.4 3 4 5.2 2.7Brazil 2.8 2 9 4.5 0)3
Forniet Sovict Llnioni 2.0 2 4.7 0.2
Indonitsia 2.6. 27 3. 0.8(ColoiNia t1-5 1. 2.4 1.)
( Oet (ivoire 1 7 .2 7 7 5.9Thailand 1 .2 1.2 1.7 0.4
Ntexico i. 6 0 6 I C) 0.1Argenitiina (0.6 (O.7 1.1 0.2
.t?ilii(tia 0. 1 o. 2 0.4 30.4
CSI '. I StilMLiw Iht (O inni1djitV 1tiliVy tll \lAilsSiS I Iii, Inicriltional tcon(niics
D)cparitmcnlt. Wtorld Biank
Managing booms and busts
(ommmodlity boomis andit hLusts create two types of' prohlems: those aris-im,g fr-omil llIct1ati S in io iC e and llth IIOSC Inrlated tO thIe faCt that fln(--tuLatiolns ill ilicoifle take the lcorm of' chaniges in the s pupl) Of foreigIlexchange. e'lic role of policymiakers is to devise ways to smiooth out theimpl act oI such iloliCle flutCLations onl the economy
'I'li lolig-ter-ili effects of0cot)inlmodity 10011mS are niot always good. 'I hisparadox is ofteni knowil as IutCh (diSese.tt " )`Urilg tile 1 960s Dutchmallufalturing suft'fe-ed from the apppreciationi of the real exchanlge ratewhein natural gas exports boomed following discoveries in the NorthSea. Conimiodity boomis anid husts can also cause other prohliels, sucIhas misallocation of1 i lvstilIeIlt resouLrces and fin ancial difficulties forexporlters. E'xporter-s' financial prohleis OCCUir whtenl they sell forward atfixecd prices comimiiodities that thyv do not own, If prices rise sharply(ahove the forward sales price), exporters Inist make Lip the differenecc.
A buoyalIt expor-t commodity in one sector otten leads to declinies ilnoniboominig export sectors. Resources flow into the boomlling sector,while investmenits elsewhere declinie. 'hTis coilin modity booms teiid tioshift profitability am11on0g tradablcs as resources shift to the boomingbusiness. ['hey cani also shift profitabilitv fromil tradables to nonitradablessince somrIe of the incremental foreigi exchliangte earnings froll the comI-moditv boomi will be spent on nontraded goods. For this to happell, for-eign exchange eariniilgs need to be sol(d ftor local cLIrrency. If' theexchanige rate is floating, the cuIrenIcy will appreciate; if it is fixed, thiemonley supply Will rise anid local prices with it anid helice the cuirenicvappreciates.
Olvervaltmation of the currency can persist for som))e time atter theboom endts. Past comimilodity hooms have bceen followed by sharp andprolonigetd price declinies. Th'l overall result 'T'he profitability of theonce-boominiig export comimoditv is severely reduced, while othelexport subsectoi-s have heenl devastated by thve large real appreciation ofthe currency.
Coin mi ocditv hoonins can also affect the gover-nlillelnt's budget deficit.Wheln export coi iiiiio(litv prices soar, goverinimen'clt revenlUeS usuallViicr-ease sharpIv. lI'is occurs directly if the resori-rcs are owiled by thepublic sector, isl s com o ill petioluli and mininig and minierals, or
8
NMANA( .IN( Bo S)NII -AMI lit IS P,
indirectly, tFrougIl increased export and imiport taxes andt hligiler incomlietax r1'evenueS. h'let boost in revellues can elIcourage governiments to com-mit to long-term and expensive programils and projects. When commiod-itv prices decline, so do governimienit revelnues, but not expeilditures. Asa result, couLtries are ofteni wor-se off than before the boomi. WitnessColombia, Kenya, anld 'Iazani.ia after coffeCe pirices boomed in the lateI 70s. In the post -oil boom01 Nigeria of the 1970s anid early I 980s. thiesho0-tl'all of rCevenLues Was so sever-e that exter-nal1l borr-owing had risenldi-amatically eveni b horc oil lr-ices lbegall to fill. It was 1muLIcIh tIle salmlestorv in Mtexico's and VCeIlaZLCIa's hadliaiigllg o)f thleir oil boomlis. Also,sinlce govel-lilnelits Leild to inlcrease their eXpenIdituLl-s with tIle win(dfalltax reveiluL, fiscal policy tenids to be pro(Aclical anid helice causes insta-bility in the economy. I Ihis, according to ( uddington, hlas a detrimentalimpact on long-term econloiilic growth ((Luddingtoll 1988).
For many1V (coimod(iitV-expor-tillg couLInties, boirowing in global mar-kets is not an option. Ior them, internationial loans teind to be availablewhtenl t(hey are inot nteeded (whein coini modity priices are high) andun.available wlhenl they are needed( (w'heln pr-ices aire low).
['rade poli(ies can also partiv deteriniie the effects otf a commodityboom. If they p revciet revenues fromil being spent On imports, traderestrictionis miay lead to further- appreciation of the exchanlge rate. Suchrestrictions inay also skew incomlie distr-ibtionll. In Kenyla after tile197(6_') cotffee bom, import and capitil controls increased reiits tosuppliers oft capital goods an(d imported colIsumler goods so that muchof the gain from the boom went to urban dwellers, who are the suppli-ers of the goods, aild inot to farmers.
In several commodity-depenldent countries, paarticularly in Africa, sig-iliticant antiexport bias already exists because of milacroecoilomic iinsta-bility, import protectioll, aild aid inflows. Ilherefore, avoidinlg the Dutchdisease is especially important for these couLltries.
Learning from experience
Governimlenits have takenl various measures to stabilize revenues, neu-tralize the impact of foreigin exchianige initlows, and avoid the onset of
)utchl disease. 'Ilhev have adopted pruideiui fiscal policies, liberalizedimport restrictions and removed capital cont iols, tightened molnetarypolicies, accum-ulated foreigin reserves, anldt redLuced foreigin debt. Butwhat were the specific measures, andl how effective were they?
FXpX)11 ,lXs. In the latest coffee boom11, Cameroon, Cote dclvoire,Honduras, Uganda, Viet Nam, and oithers reintroduced export taxes.SucIh taxes, so the argumenlt goes, can alleviate both the spending andthe resouLrce mnoveIeneit effects of the boom11 and protect nonboomiligtradable sectors. lax revenLIes shOuL1( be invested in high-yieldilig pub-lic investmiienit programs that increase long-termil growth.
In practice, however, tax rates are usually set too higih, and publicinvestilmenit projects are often selected oni nonieconomic grounids. rhistends to damiage the boomiiilng sector with little compensation elsewherein the econoiyv. Andi there is the general risk of larger future deficits ifthe goverimeint makes long-terimi co ini-nitnienits based on boom -tinmetax reveniues.
Excessivel, high export taxes can cause other pr-oblemils, suchi as taxevasion anlid smugglilng, that call lower tax revelnues and cause seriousdisruptionis in the domestic market for the booming comimlodity.Llganda imposed an export tax of 32 percent on coffee in November1994. Some exporters did not pay the tax and, as a result, were able topay growers highier prices. Exporters who paid the tax (but still had tomatch the tax-dodgers' farm-gate price) are now, apparently, in financialtrouble, Iow to moderate export taxes (acting as a windfall profit tax),however, could have beneficial effects by reducilng overinvestineit in thebooming sector and providing incenitives for diversificationi. Civen the
actual and potential problems, policvmakers should be careful whenusing export taxes to manlage cominoditv boomns and busts, particularlywheni setting the level of the taxes.
(Coninodiry booim lbonils. In thie late 1970s and the imnid-1980s coffeebonds wei-e issuLed bv the Colombian central bank to private coffheexporters or the Federation of Coffee Growers in exchanige for dollar-denominiated bonds as partial paymilenit for export sales of coffee. 'Ihe
1W
I I,\RNING I R() I \P SI'I No I I I
coffee boInds were palrtiCUlarly important foI alterinlg the monetaryimpact of tile coffee boomn. Thile bonds acted as Iforced savinigs in foreiginexchange kept at tilh central bank and sterilized foreigin exchanigeinflows v reduciLig their impact oin the mnonietarAy base. An advantage ofSuIch bonids is that they are usually more readily acceptable to the privatesector thani exxport taxes. I lowever, boom bonds often involve somile
export taxatit In ill disguis. (Certaiiily this was tr-ue in Colombia in the
late I) 70s, wvhen the goverinimlenlt negotiated with the Coffee GrowersFederatioln to get its memhbers to accept bondls hbearing less than tIle CUr-
renit market late of inteL-est-an1 inlplicit tax.
Ihle honds wouild have had difterelit (oonse(lueices if they had beenlissuLed by thlie govei-rinmenit ratiler- thiani tiet central lbanik '[he goveriin mentimiighit have spent the revelnlues f-r-om11 thIe sales of thle bolnIs, Which woUld
have nega.ted the steriliz ationi and national SaVilngs eff(ets. In other
words, if- tlih goxmernmiet spenlds bond revenues, thie effect is siti lar tothat of exporters Splending Wilndfall revenues.
eilvisitt l)ll. ( Coverniiments call proimo(te diversification by tsiniginceal-CsedCl r-evFene's to enhan1,vCe exteIsioll and research services and to
finanIce in-I.rstrULCin Ire prOjecls. D)iver-silik.ation of export com mod itiesshOuld, ill general, contlihute to the alleviationl of lutch1 disease. (sing
sO111e of the ilnCotlineC I-0oIl highl Coffee prices ill the ear-ly andl id- C '8)s,
tIlh (0olomianill IC'ederation1 of Coffee ( roWCIs laullChed a diversification
programin in (offee glowing areas. 'lie I'cderatiOn mundeI-took resea rch onil-t-rnative6 Cl-ops thlat coIi Id be produced ill co-f fe-growvi ig aieas and plo-
videdL COtflf'' f<l-irmeIS Witil eteCnsionl SeIrvcS. I 01 1ilWoIr-ilnComelC COLIntrieS
the lesson of diversification is that fLliliilg to attach a suffiCienlYlV hiigl plri-
ority to agrI-cultr,1e wVhile fariing Othle settor-s is CoLunterproductive.
ManytiVf these Coulntries uln dertook ind-ustrialiZatioll projects to diVersify
tileir- eConIOi iieS withlolt giVing duetI (onsideti3on to physical and huILmIan
infrastr-uctu-e. lhe exprCl-ielce of rapidly glowing Asiall ecollomilies indi-
cates that a tLIvUtnliC ag ricultur11a1dl C10O I(oistiltties ll imploi)tlailt ingrldi-
enlt for su(-cessful diveI-sific-ationI a nd, eve1tu1ally,. industrialization.I h'd'{i,nl I iIedgiilg ill( n t-easS tIeC plreddiCtability Of fri tillre revenILies,
rdtedl(ces u in rtYlilitV, ati)(d M iakes p)lari n uig ts sler. I ledginig cani also assistgovervinieritis and t'he plrivate sectolr to siMooth short-terCm-il price fluctua-
tions and a julst to new Comi0iio0dity prlie t treids. ( ovenrinleiits ad tilthe
prisate sectlor (mi iron out fluctuations ill future rVeIlUS bv uSilng
optiolS, Swaps, 1and Iltu res 'llt The Meta 111 g(OV'leiliCHt uIse SLIChI h1ecig-
inig iocs to CleIsm-ct guariitee(ld pric OIf- its oil (aid(l predictable rev-
eInueS) after the ( ulf war. At that timle, oll prlicC ULtnCeltaiInLy and volatil-
ity were highi. (ne of the fiears was that pri(s mlighit collapse if the warenided quit(kly-wh ich it did. letween [)ecemnber 1 99'( and l:eriluarv1991, throulgh the USe of a colImplex nlliX (f Oil futures, options, andSWapS, MNexico hedged aboUt SiX m )iIt iS of oil exports. It was thuLseCisLired of a fixed (arid knowin) reVenLu, illowing poolicvmaikers timie foradjUStImIt wiein pri(ces djl(I collapse.
12 M N %(;IN(. ()MOI FN IMMAS -A`KI) lltSIS
Governtilenits can also use risk management tools to protect prices andtstabilize the revenues of small holders, who operate at volumes that aretoo small to make it attractive to directly applv risk maniagement i nstru-mients. For examiiple, consider a country suLIcI as Mexico, which offeis aguaranteed mlinimlium price to cotton grower-s prior to planting. [ivoffering the guarantee, the goverinmenit has assumlied on behallf of thefarmer the risk that cotton prices will fall below that minilum.lil Inmany countries with minlilmlumLI price guarantes, the treasury ultimilate-ly bears the cost of such programs. In Mexico, however, a governimlentagency puirchiases options in the international market, transfeirinig therisk off the governmenit books. 'I''he cost ot the option is known whilenthe guaranitee is issued and remains tixed evenI thougIh the eventual cot-ton price is unlicertain. In Mexico the c-ost of the program is chargedback to the participating cotton farmers, with little cost to the govelr-ment. '[he [IInite(d States (for grains) and Canada (for cattle) are exper-imenting with fuituies and options to provide protectioni to the ir farm-ers. 'ihesc instrumnents are usually Iess costly than traditional policyintervenitioin. Government offers of risk man.agemnent instruments canbe partiCUlrlly useful dclrinig periods of tranisitionl troin governmaeiitcoltrol to m(arkets, befor-e viable private sector risk sharinig anld riskmnll"agemnent arraInMgements hIavC had timelC to developi.
'I raders and prodLiceFs, too, canI benIefit from hedgiig. 'I'hev stand toinCuL big losses if the price of a comimioidity rises sharply and Illexplect-edlv after they have comimlitted to deliver the conmmodity\ whichi they donlot yet owIn, at somie fixed (and lower) price. Ixportel-s would losebec(atise they wouldl have to pay miLCh highier prices to farm-ilers than theconililittedl price. Similarly, trader-s inlight contract to bluy the coinmodi-tv Ifromi farl-mers at a fixed price in the fL,tUre but would suffer if the pritcfell. They would receive a lowel price Whlell selling the Comm1lloditLy, andthat price mligILt not cover the higher price they paid to farmers.Indonesian coffee exporters are said to have gonie bankrupt or Suffered
mammoth losses in the miid-1 980)s (and in this past yvear) becaLIse thevdid not take appropriate hedging actions. I Igandan coffee traders facesimiilar- pr )blemns.
Market-based risk management instIumeniits iare good for dealinig withirelatively short-lived price disruptiolIs, bcLuse of- their maturities andmarket liqluidity. Agriicultural commodities are a case in poilt. Maturitiesand liquidity for metals and energy are longer, making longer-term hedg-ing feasible.
I ledginlg involves certain costs. 'I'he USe' of LItLures involves miargini pay-menilts that at tinies could create cashi-flow proiblems. Options involvethe paVInCI)t of a preinium (if bougit), which is the cost of insuranlceagainst price declinIes (put Options) or increases (call options) ManycouLItries ltce creditwolrthiness problenms in ac-cessing swaps and comn-modity-linked bonids. At timaes there is negative plurblicity associated withthe use of hedlging instriumeIts for example, locking in a price for
I IAIINI\( I RO I \11 1,111 N( I I;
exports and theln witinessinig a price rise. Also, hedginlg oftcn requiresestablishiing a heciginig programil, such as installing appropriate accouHnt-ing and moniitorilg systems and traininig people to trade. T'hse costsshould be compared with the- bLelnfits [tom1 reduiccd uncertainitv.
I low muLIChl ShOUld a couLitr or a private firimi hedge? '[his depenids ona numibher of factors. Amionig the miost important are the identification ofthe risks of not hetdginig, the risk aversion of the tirm (or goverinment),the volatility in the market and so on
l inally, hedCtgilng that uscs mar-ket-based linancial instrumIenIts adldress-eS price Uncertai ntv, andl mIay not necessarily stabilize IrVen1Ue'S. I loweVer,in miost cases the volatility miaiilyv coImtes Iro nIl pr-ices rathler thanl froIllouLput. In this case, these instrumnIIlts assist in stabilizing reveIues.
One( caVCat: hCdgilln illStr-LuImeiCts nl'eed to he traded by professionalsaniid with a clear hedging pUrpose. If ltismIanaged or used for outrighitspecul ation, theC c(anI caUseC as mlLichi or Imlore damilage (because of lever-age) as unliedged ilovciicits in prices. l'hIus the usc of hedging i nstril-menits by countries with poor cr-edit rat i ngs and wcak adm inistrativecaapacities is problematic.
A( Ccess l otialncial omarkets. In mallny coonn ies thiere ar-c few alicttl-ativesto the hoom ilIg sector for investIlmcit Liberalizing and developingfinancial markets can provide option's for alternative investlienilts.HIldcdrdeveloped (lonlestic financial and(l capital markets ilma nlot bemuLcIh of a proHlem, however, if producciL scat invest frecly ahroad.
( C011fiting'emm lwrlmovinIIfIcilictis N4aln collmlmodity-exportillg countrieshalve little or no access to international lenders wht]l theyc nieed itmost that is, whcni prices have nose-dived. Contingent borrowing facil-ities, sLtch as thc ( olpensatorV and Cooti ngctncyv inancinog l'acilits ofthe I VI F and thle I trl-opeall 1Il ion's S'1!\BIlX, ac (lesigined mnainil to dealwith tenMpola1tV Shortfalls in cornrnodit} export reCVClUeCS aftrtl a bootn.'I'lhese programs p 'OVide It'lidilng after a COuIntr is idcntified as havinig
suffereud substantial expoIt CVCIeLet shoI-tialls.Rev'emtne simlilizaliomm Ifiinds. AInotlher- way to manage con]malodity lrice
shocks is to create a stabilizationl fUnd-- not to support or stabilize cOIll-modity prices bLit to provide a guide' fol stahilizing govertnment expcil-dituel-s. (ile le has suchl a ilund for coppei (box 2) lly savitng resouLrces ill
hoom years, govern-lmenits can reduce the Ctied to CUt spenidinig in leanertitmues. A we.ll-desigtned fuLnd can stitooti expenditutre adjustmllentsUlpWards or- downwards. Rescivs in stabilizatio0 fulIds should be treat-cd as intel-national foIeigni reserves.
Stabilization ltittds are like collmnodity boom] bonds, ster-ilizing for-Cigil eXChange1 in]flows tesu-ltil rg Ironl a bo om I'Lli unds may et]cotlittt- dif-ficulties, however-. Price slHiII)s teind to last ninch loiIger- that] booIms, soto achieve stabilization, large reserves i]e ed to be deposited into thef[1ind, and that colIId imiply financial (clsts in teritis of forgone opporto-i]itics lor othel- piolnitable i nvestrmletmts. A large [m in( is also often not fea-sible fIor domiestic- pl itical reasonis, beCatIse of spendinig pressurcs from
I-I l\\ \(.j\t ()\IN1(tl)I IX IIW( S-AN I)tI)S lis
domiestic coInstitueLicies. Anld even with lar-ge deposits, thtre is still thepossibilitv that a f'undi may go bust. 'I'le viability of such fLunlds depenidslargelY on1 h1ow long the cornmmodity price chanige lasts and on the stabi-lizatioll rIle aidopLted. If exogenous shocks hiave a persistent effect Onlprices (or Prices revert to thieir nmeanl slowly), the financial sounldiness ofthe stabilization f-unld will ultimately come illtO queCStiOn. StabilizationfunLdIs coLlIt also be subject to In isinililaaelilenlt of thleil- resources-somewhat sinmila r to problems encotlinteletl with export taxes. I lowever,this is nlot as gorat a (danger as in the case of tax r-'VenlUS, SilnCe Stabi-lizationi thtd I-nd rveuIes al-e 1not included ill genlUR-. govern ment revenutes'I'he fulid also ptrovides clear- rules oll whenll these revnILues should bespent, which is not the case with export taxes. Stabilization funds Ima.lV
be usefiul whelnl dlomiestic fillancial miarkets are not well developed, andthere alt ic borowinwig consatrailltS. AISO, a fillt( nMa be useful whe1n aneco110nov bears SigIlifiCallt costs When it adldusts spending in responise to
cornlmt)ditv booi l11 and husts.
I lit woiki ngs of a stabilization InLrid call he cornplleieinted withi mar-ket-based hedginlg nstr-u inenltS, suCh .IS ILitLnres, OptionlS, anld swaps.
Market-bIsed hiedginig instl-rlitilts art ideal for redlucilg thIe exp)osuLe tocomilimiodity prliCC lHtcetaintv, ald a SltahiliZatiOn fund can redUCe the
instab ilitv in goverrlmllit i-eveLtLies (ot expenlditrL ys) caused by corin-)d ilty, priCC chalngeS. Coinbilli ng te Stabili ization) funldc with hledging
reduces the probability that the stibilihzition lundld will rill out ofreserves ,nid allows the fLind( to be significni 1vstl smller.
I lowever, it shouild bc noted that the ilntCHrn 1tri 1 nl exper`ieneC withl Sta-
bilizatioll fun1ds is that ImOst of the ltil ilito financ11(ial difficUlties. I'tis
p )I icrmanAcrs sh ould he partticularly ( alit ious when decid inig to ad pt sucIa scheme.
ftit('tlt(ltnttit , <tottittodllVi aVeettt(Yo l).,i ing tIlh late 1970s a11nd tIleI')X(s n1iarty eolilnm1oditv-priodtcLCilmg (cL11itrieS attellipted to stabilizeCminiiodlitv p li(Les tIrl-Ingh ilnterlnlati01l0ll lgla I l1remets. Ihese ilCInelttCd thlt
Box 2. Chile's copper stabilization fund
Chile's copper stabilization fund was established in 1985 as part of the gov-erminenit's structural adjtistlllelt program with the World Bank. The funduses foreign exchanige reserves to flattenl cyclical variations in revenues. Tlegovernmenit sets a reference price for copper (related to some moving aver-age of internationial copper prices). When the actual price rises above the ref-erence price, a portion of the excess revenue received by the government isdeposited into the fund. The government can withdraw funds whell themarket price is below the reference price. '[he find's reserves are treated asinternationial reserves and are kept by the central bank in an offshoreaccount. When copper prices were high in the late 1980s, accumulatedreserves in the fLind were used to buy back some of Chile's external debt.
I I ARNIN\t I RO' I \1'1 R1IIN( I I )
Interinational Coffee Agreemelil, which uised anl export quiota system,and the International Cocoa Agreement and the International NatUralRubber- Agreemient, whicil used buffer stocks. Only the rubber agree-menlt is still in force-just barely. Its effectiveness is being questioned hysonie signatolies.
In January I'M95 soIe coffee-exporting coulItries introducedI a stockretelitioln schemelil to peg highi world colffee prices. Exporters are tostockpile a percentage of exportable pyrodiuction (that is, pIroductioninlilnus domestic COnISUIllptiOn). 'I'his SChemeIl> Will be difficult to manin-taili, hoxwevel-, evell ifiall expolrLtil ,countries joini thle prograill. (Ifoonlysome exporters joiii, the free-rider problem arises witlh counltries that
benefit from the scIelmil bLut that dIo not participate in its costs. In thecase of coffe(, the costs woulld be adminiistrative andl storage costs.) Ifworicl prices are kept artificially high, Irloduction1 woulc likelyinlCrease S ut WOLdld COUIltr-ies be willing to act umllIlate ever-increasingstocks!
(Cocoa exporters, too, have recenitly beeil discussinig plans to cut prro-duction-by 375,000 tons over the next iive years-or to stockpilecocoa to hoost (and to somie extenit stahilize) world prices. As with anyothler commIIIodlity agreillemnt, there is thie free-rider problem, whlichi is
likelv' to undermine ilhe scheme, More general ly, experience with price
Box 3. Colombia's tale of two booms
In the past 20 years, Colombia has seen three major coffee booms-in1976-80, in 1986, and again in 1994. The first two booms were handled indifferent ways. In the 1976-80 boom, inappropriate fiscal and monetarypolicies meant that government expeniditures began expanding rapidly in1977 and accelerated in 1978 after prices peaked. Most of the increases werein governnenit consumption. Neanwhile, revenues grew modestly and thefiscal deficit expanded, financed by increased foreign borrowing. With noreductioll in government net credit to offset the large build up of foreignreserves, the monetary base expanded. The exchange rate appreciated in realterms by 30 percent between 1975 and 1982. Noncoffee exports declinedfrom 7.7 percent of GDP in 1 976 to 4.3 percent by 1983, completely revers-ing the diversification efforts in 1967-74.
In the 1986 boom, the government responded with less expansionary fis-cal and monetary policies. Most of the increased revenues were held by theNational Coffee Fund. Some were invested in dollar-denominated instru-nients, sonie used to purchase bonds forni the central bank, and sonie torepay external debt. The net effect was to sterilize about 60 percent of thewindfall revenues though external debt repayments and about 20 percentthrough open-market operations. Only 20 percent of the increased revenuesentered the money supply. On the fiscal side, coffee tax reveiiues were usedto turn a deficit of 5.2 percent of CDP in 1984 into a small surplus in 1986.
]6 \t 'I..N ( AN \INt( )III Ito()O\ l5- DNI)I HhISI
stabilization programs (buffer stocks anid export qIuota1s) suggests thatmarket interveition is likely to fail in thei long rul. Whenl that happens,comm111odity prices usually plunlge. Witiness coffee pr-ices after luly 1989,wheln coffie-expor-tillg couLntries starteLd to rielease the large stocks theyhad acco mulau3teCd WhenlI expOI-t quotaLS we're, in for-ce.
Fist-icl dis-ciplillc iil resspom? to comm iolotit ' pi ire bwoomns. Receint resca ic h ispoinitinig to the strIcturLe of the budgetary process as a kev deteriimianitof the nIature of the responIse to termiis of trade windfalls. One problemis the existence of multiple claimants to the increase ini revenues. Ilhese
could he powerful miniistries, initerest groups, and state-owined enter-prises. M,ultipIl claimanllts mayV appropriate molre thalil the elitire pool of1reveLues, divertillg it to less-productive activities that beniefit their coIn-stitLIelcies. AlIthouigh it is in the interest of all claimants niot to stflanderthe revell) e pool, i ndividnlal ckain man ts h1ave n1o inIcetive to act pru-dntIlIv if- ther-e are nIo gual-rantes that other clainiaiits will do likewise.IhlliS lil ' ofr-CSeallrC suggests that prHde(nt ma n11agemenIt of coilillodiitVprice hot0nIus could include chaliginlg budgetaly priocedul-es to make itless likelN that fiscal wincifalIs will be squaIldered. Greater centializationlof control over budget procedures migilt be one wav to hol down thenuhmber of claimailnts to windl 1I1 reveLues. If the p roposed solutioll is arevenlue stabilizatioll fuld, the anIalVsis highIligilts the importance ofStrictIa isolatilng the suLI Luses accm0111-ate I inl the fun11d from-1 the genler-
pl1ool Of reveuLIes. Resear-ch also shows that coomimiodity-depenidentiCo u1tries often stUffer fr1om Se\ver'e teHIrS of trade shlocks, and this, intLirt), has had detrimental eflfects on1 their long-termi ecoInotIiic growthalnld inivestililet.
Who manages the boom?
At first glanlc, it seems tihat any booim migilt be milore easily ilmanlaged-and tile impact of Dlutch disease softenied-ifi comimloditv prices wereconitrolled by gover-nimienit (througIl a miarketinig board, for instance).'I[lh govelrlillnmint nteed not raise taxes hecause it couldi adminiistrativelymainitaini fixed iterinal pi-ices. Lvidenice suggests, however, that thliscould cauIse Imlore seriOuIS pi-oblemils otf adjustment. Fxperience inColonihia and(l francophoric Suh-Saharan Africa shows that of-ficiallv setprodul(cer- prices are niot inisenisitive to wonid imiarket condii0ious. Wheln aworld co0nim lditV pr-iCt ilncreases sharlyfv ( (r)offee, fOr exaIIple) thereis pressuL-e to raise t'doimestic produceCr pr-ices \When the boom endlts,hovever, goverinien iits f-imid it difficult to bOWer- proFdlutCer pr-iCeS. Indeed'd,goverinmenits often subsidize com1 o1dity prl-ic- after theI boo)m to mlakeuip the difference between pr-oducer- pr1i(c(s and export pr-ices, leading toSeriOUS fiscal problIes. '[hat happened in ( (jte dfvoire in the late 1 980swheln world cOCOa priCCS declilnedl alId il ( oh0ombia in thc carly l )(Oswhe n world coffet prices decli ned.
Some econiomilists argue that booms anid (espeCially) busts that areobviousIV teHIpor--y are tIar fi-rom1 typical Iore Comml(non are the longand unlforese(n colin miodity price sltiLmps, like those of- tie late 1980sanid the I )90)s. Even wheln the boomis are( orrcctly perceived as tempo-rary (the oil price boom in 1 )90-)1, for inistanoce), goverlillmelits imi imle-
dliately spend(i thie windfall revelues. Look at Niger-ia, where the goven-ll-niInlt spelnt (ulickly and, fi-omil the perspective of long-termi growth,poorly.
So, if the track r-ecor-d of governmienits in managing boomis is far fi-omilimipressive, often becLuse of veak administrative capacity, whiat aboutfarmersls? Thller-e are argulIlelIts in favor of leavinlg at least part of revenLueStabiliZatioln to farmer10-S. Much of the dehate lii nges on1 whethel-r farmerscia niake the right choiccse I hle eviden( , while far fi-oni compiplete, isfavorable. [)urinig booim times, somile fi'rIrier-s have used the windfallgainls in a rational manner. Kienyani coffce allrime'-s uin derstood thei teml-po(rary n1atul-( of the coffee price boom in thie late I 970s and savedi about60 perment of tleir- extrla incomie. Rice faimers in TIhailand smioothi theirConsLimptioll hoth withinl and betweein harvest years. It is not necessarilvtrue, therefo,e that in the absemice of govermn mit stabilization plogramils
I 1
1 8 V ANA(,IN(. ()INIODII)IY BO()MIS-\AN[) HllISI\
far-mllels would overplant in response to teiimporarily hiigh prices, so longas ther-e are alternativc economic activities openl to them.
'[he p ruLdenit behavior of farmers, however, may still be frustrated by[)utchl disease or by an othel-wise inappropriate macroeconomic enxvi-ronilienit (limited access to foreigni exchange, for instaice). Controls onimports anid restrictionis on investing abroad miake this all the more like-lv. '[he implications for policy are that governmilienits should allow theprivate sector- to play a bigger p art ill stabili ilng liuctuations in corm-modity-rclated inicomIle. 'I[lat is likely to be et'fective, however, only if-othel- reforms on1 market and filiaiicial liheralization are uLl(lertakensii mu Ilt a(ouslx.
Alternative policies f'or copinig with the problemiis of commoditybooms, and with price volatility more generally, generate different setsof winniers. For examiFple, export taxes instituted durinig boom years willpermanently transfer large sums from low-inicome farmers and the rulalsector to the government. Alternativelv, if farmers hold the windfallgains as Savings in foreign currencvy bonds, miost of the brenefits fromil thebooml iremain with themii.
How to manage booms
Sinice the receint price increases are expected to be short-lived, develop-ing countries should manage the bulimper revenues carefully. Fxperiencehas showin that couLItries tihat adopted prudenit fiscal and monetary poli-cies, accumulaited reserves, reduced foreigin debt, anid liberalized capitalcontrols and import restrictionis saw long-terimi beneficial results fromcommodity boomns. Other, complementary policies have also been sug-
gested. These policies cotld provide additionial benlefits, depenlding onthe conditionis of the economies concerned and7 the nature of the boom.Policies employed to maniage commodity wind(f'all revelLues are:
DLon't ov'erspend or overcomnunit. Governi menits should not commit tolong-term spenidinlg that will be unsustainable when the boom fizzles.'I'he experience of the 1970s shows that countries that adopted aggres-sive spendinig programs expecting the high prices to be permanenit facedhiuge debt-servicing problems when prices (and revenues) fell. To pre-vent this, a large slice of the incremiienltal reveilues should be used tobuild up intterinationial reserves, to be used whenl the boom ends.
Adiopt prudent nmonet(airy anld lisc-ail policies. (oninmodity-dependent coun-
tries are more vulnerable to external shiocks, making soulid macroeco-nomic policies especially critical. In this context, monetary and fiscalpolicies shotild not be expansionary, and countries could even considerreducilng exterinal debt with the proceeds l-romil thle boom. Recent researchshows that comimiodity-depenident countries often suffer from low eco-nomic growth because of wide fluctuationis in thieir terms of trade.
Monetary policy should balance concerins over inflationi with coIn-cerns over currenicy appreciation. Capital-account controls (restric-tions on investinig abroad) and import controls should be reduced toprovide a way for foreign exchanige to flow out without affecting themonetary base. Investinig part of the boom proceeds in foreign assetswill sterilize foreigil exchanige inflows and avoid expansion of themonietary base. Simiiarly, comimioditv hoom bonds can be viewed assavings in l1-0reigil currency and as aniothier effective way to sterilizeforeign exchange inflows.
Consider Iradc lileralization. Govern ments slhould consider taking
advanitage of' the favorable trade balanice situation to liberalize importrestrictions. 'I'is will hold down real exchange appreciation and
I')
20} \I \N V\IN(. ( ( )\NI()I)I I IOS-ANNI'NI , I ,ISIS
iniprove economic efficiincy. If Willndfalls are used to finance triade lih-erallizatioll, thelr- will be permanile incomie gainis and trade diversifica-tiot, and the econoiiv will be in IIiiciI better shape after the boom thanit was belore.
I)ii'Crsi/t l(it ad ljuus. (,>vernlenll tS coul<l facilitate diversificationi hvuSilng inem iliental revelLues to enilalice extenision and research servicesand fIlilnic inIfrastILuctuLe pirojects. Along with diver-sificatiol, structuralad jistinen t put granil s shotiutl he ilin p1 cileltemd to illcreses economic efi-i elnylV palrtiCHlalyA ill tile boollilg sectors. It is Iss pailiif-ul to begin
such progr-am1s whenI pI_rices are high alid se(tors will he better able tocole VVith IONNTI- C(lllloditV prices w ilC booni sLhSidcs.
A slabiliziliiu imo? 111t1 hl' bct opt iOi 11tMICo't %tot11iCtitttiŽ tiS By savingrevenuLIes in hoom years, governiiilets Call iron oLit explendituLe adjust-ments upward Or downward anaid IrteLduce tIlh nleed to CLIt spelldilng ill
leaier tineIs. I 'ie fli-dLIs shortld be treated as inIter-latiollal for-eigilreser-ves. I loweve, fritids otten unll iiito financial diflictilties because of-the in1pl-edictability of co0ni1modity pirice b'oelhavior- Itlhy ar also subjectto in isma nageinen t
11c,gc. X Market-hased hedginig i nstruLilmeritS (futuUres, octions, swaps)call incr-ease the predictibility of anticipated coinmodity-l-elated rev-enues anid make planinlilng easier for foli(yinakers. 'I'hey canl also assistprivate sector exporters and traders in handling their commilodity piricerisks. Wheln coinplenientirig a stabili/ation fit fd, thesei nIstrulmlelits caniiiron ouit flows into and out of the fuLnId. lecaLuse hedginlg is a fol-rll ofprice insurance, it coiiies at a cost. I lenieiits of thlis cost inciude theoptions preniiutii the op ontuniity cost of ial-gill accouLIts, anid theadmiiiiiistrative costs for settiliig rIp aiid rIinii'ig a hedginig program.I''iese costs should he compared with the hbenefits of hedging-a reduc-tioni in uncertainty and a smiootiling ol sh ort-teriil price IirctLiatiols.01nlV COUlltrieS vith good credit ratings and aiprtloriate adminiistrativecapacity shltlIld cotisiderC rISilng Ilidgilig illStl-ruliiients.
Avoid puniitilc exporIt e I(IXtSI ligli taxes nIalv e11Co( rage tax evasioni aiidsniugglinig. Moreover; govel-lilielits telih to spend increeilelital tax rev-ennues swiftly and Apoorly. Aiid expoort taxes caii only offset foreignexchalne inifows if the tax proceeds are hfeld in foreigii assets. I lowevei-,soine level of taxatiinll could redrIce overilIvestlleClt in the booming sec-tor anid firovide all incentive for diversification.
A void o tc -tsh tv t dot twoot. liite ni-atioii aI agr-ee imeiits (oil bri-ffer
stocks arid export 1tiuotas) are uLilikely to keep commodity prices highalld staile in loIng rr-ll. In the past, iiialy coiiiliodity-exportinig coulitrieshave entered srucli agreemetits ill ai attempit to stabilize aiid, in iiaiivcases, raise prices. Inl aIlythinIg hut the sholrt teriii, iioiie has suicceeded.
References
Adams, Robini (. (RCsOLIl(C StrlaltegieS, In ). 1I9951 I1tervieW ill 'TcI itii1ialc i n Flllnk't, March .31
IlLICI-, Pete r. 184. "l'mhircilli-elc of1Studies last: Retracing First Ste ps."In tC-lld NI NIt. eic a dl D)LndlIC Scels, c(s., IPiouicee% in I )(,''c1lopllh1l'nNeV Y0ork: ()xhr0ld Ilnive-sSitV Press.
(hlCsSCIs, S., ald R. (C. DuLnIca. 1 99 3. Manp,nia'o ( ot,unu'd,t ltMice M?A', in
Dlci'elpin, ( .onIitri' Baltimoe:i I'li lohns I lopkins tliniver-sity Prtssfor tile World Bank.
Corden, W. NI. 184. "Boomt i ng Sector and D)uitch Diseasc Iconomllics:S'LlIAT a111id (ConIsolidat ioll." (4)xl-frd 1.,-tomimi Papcrs 36: 359-80.
(Cuddington, 1. 11988. "I iscal l'olicv in (Comnimodity-Exportinog LDCs."Policy Research Working Paper 33. World Rank, Washinlgtoni, D.C.
Davis, jeffrt NM. 19')83. "'Ihe' I:co<noioic Effects of Windfall Grains inI xport Fari-ings. 1975-78." D\4uh ip I) itnclillopn 11: 11i9-39
Deaton, A. "(Conmoclity Prices, Stabilization, and Growth in Africa.[)iSCnLsSioll PI)aper 166. PIrincetoni lilnivei-sitv, Woodrow Wilson
School, Research P'rogr-amil in Developillent Studies, P'rincetoni, N.J
ldwards, S. 1984 "toffee, Nloiiev, and Inflation in Colollmbia." WvAorldl)D,clopuucmt 12 (11/12): 1107-17.
0i bert, I_ ( . 1 994. "1 omilloditv I'und Activitv and the Worldl (CocoaMarket." Lorncon ( Corlmoditv lFxchangc.
llill, C. B. 1')'91. "Managing (Commoditv RAooms in Botswana. Worlvic)pt't'loplu('ut 1') (9): 1185-96.
Hlill l'olly. 1963. '11i Mlijg'lam ( ]ocoa Faomcis of Sintherti (Ghana: A Stni'(,lini Rual, ( apitali,sm. (Cambrldge: (ambridge l.Iniversitv Press.
Killick, 'l'onv. 1 984 "Kenya, 1'975-81 ." III lonv Killick, ed., 'I/c lMIc lWan(Stabilization. 1 orl)doll : I leneltalll1.
little, 1 NI. 1)., R. N. (ooper, W. NI. Cordell, and S. Rijapatirana. 1995.Boom.l ( 3tC,5, all'i Adjtistmiit'i: -Il/u Xlacro' f-e XC,llol I;xIC'TOeC' ot )evlopingi
(Commi('.. New York: Oxford I liiiversitv Press for tie Wor-d lialik.
*
\IANA( I\(,(()N1Ni()1I1[) I;()(),OMS-NI) DIISI S
Paxsoni, Christina H. 1992. "Ulsing WeathIer Variabilitv to Estimate theResponise of Savings to Iransitory Inicome in Ihailand." AelricaiLcomonilc Retieu' 82 1 5- 33.
Scherr, S 1 1989. "Agriculture in anl Export tooml Econoiyiv: AComparative Analysis of Policy and Perfborimanice in Indoinesia,Mexi(o, andt Nigeria." WVorlI Developmewint 17 (4): 543-60.
4lornell, A., anid P. Lane. 1994. "Are Windfalls a Curse? A Non-Representative Agent Modlel of the Current Accoulit anid FiscalPolicy." NB11R Workinig Paper 483). National Bureau of EconomiiicResearch, Cambridge, M'vass.
World Bank. I 91). Wo)rld Devlclopment Report 1991: T1he Cltalhetigc ofDevelopiencu. New York: Oxford Il niversity Press.
_ 1994. Wolrld De)elopmIlenlt Reportt 1994: 1Iifrastrucltre ?for
LD'evkelopment. New York: Oxford II niversity P'ress.
. 1995. Comodnitdy AMIarkel's dIi thIIe De)elt)ping Countries 2 (2).Washliingtoin, D.C.
I|;luliliii l'
I ; itilli A I 'i
1t
I S-
| l | l:LE iE *1 a 1,
lE
a la J J |_I
*'T r r 018 | I
-I | i |
Ulli ;ii Si iz
| ilu, B, laI a 1|1 * | ^ M;=;
_ 181!11|31314al90 ! 1114111 1
