Economic and Distributional Impacts of Biofuels in Mali

Transcription

1 Econoc and Dstrbutonal Ipacts of Bofuls n Mal Dorothé Boccanfuso, 1 Massa Coulbaly, 2 Govnda R. Tlsna 3 and Luc Savard 4 1 st Draft - May 2010 Abstract: For a fw yars, a bofuls rac has bn obsrvd bcaus of th ncrasngly hgh cost of ol and th ncrasng portanc attachd to nvronntal protcton at th world lvl. Unfortunatly, on nots at th sa t a bond stablshd btwn th rsng prcs of foodstuffs and th dvlopnt of boful producton. Also, th dvlopnt of bofuls n Mal ust satsfy th doubl objctv of copnsatng th dffcults of ol product supply and avodng ovrly coprosng food safty bcaus of xcssv prssur on grounds that can b cultvatd for nds othr than cral producton. Ths study carrs out a thorough rvw of th prospcts of larg-scal xpanson of bofuls n th country, and ts conoc and dstrbutonal pacts. W look for th nvstnt, producton, consupton and trad of bofuls and th corrspondng fdstocks, potntal rols for bofuls consdrng country-spcfc polcs and prorts on food and nrgy supply. Spcal focus s gvn to dffrnt varts of jatropha. Econoc and dstrbutonal pact assssnt of boful tchnologs n Mal s prford wth a crosulaton CGE odl. Spcfcally, w us th topdown squntal approach gvn th crodata challngs. Th dstrbutonal analyss ncluds applcaton of povrty (FGT), nco dstrbuton (GINI) and pro-poor growth analyss. Kyword: Bofuls, agrcultur, coputabl gnral qulbru odl, cro-sulaton, dstrbutonal analyss. JEL: D 5 8, D31, I32, Q17 1 Départnt d éconoqu t GRÉDI, Unvrsté d Shrbrook, 2500, boulvard d l Unvrsté, Shrbrook, Québc, Canada, J1K 2R1, al: doroth.boccanfuso@ushrbrook.ca. 2 Rsarchr, GREAT Group d rchrch n écono applqué t théorqu, al: assa@gratal.co. 3 Snor Rsarch Econost, Th World Bank, 1818 H Strt, NW, Washngton, DC 20433, USA, tl.: , fax: , -al: gtlsna@worldbank.org 4 Départnt d éconoqu t GRÉDI, Unvrsté d Shrbrook, 2500, boulvard d l Unvrsté, Shrbrook, Québc, Canada, J1K 2R1, al: luc.savard@ushrbrook.ca.

2 1 Introducton A nubr of dvlopng countrs hav st targts for supplyng nrgy fro bofuls. For xapl, Inda plans to t 20% of ts total transportaton ful dand through bofuls by 2017 whl Brazl s plannng to xpand ts boful xports. Howvr, th food crss has gnratd so concrns rgardng th xpanson of bofuls. Although thr s no clar consnsus on th caus of th food crss, analysts suggst that th producton of bofuls s on of th ky rasons for th ncras n global food prcs (Mtchll, 2008; Hady and Fan, 2008; Bar, Clnts, Grffths and Ihrg, 2009). Morovr, th nvronntal frndlnss of bofuls (for xapl, th dgr of carbon nutralty, th pact on land us and carbon rlas, watr dplton and polluton, bodvrsty loss and ar qualty dgradaton) has also bn qustond (Doornbosch and Stnblk, 2007; Sarchngr t al. 2008; Fargon t al. 2008; OECD, 2008). In spt of ths chargs, so Sub-Saharan Afrcan countrs, such as Mal, hav xprssd ntrst n bofuls bcaus ths countrs do not hav sgnfcant land supply constrants for producng bofuls and thy ar dpndnt on portd fossl fuls for tng thr transportaton ful dand. Th nrgy consupton of Mal, n 2002, was 3.2 llon tons ol quvalnt (Mto). Ths consupton s donatd by th boass to a total valu of 81% (wood, charcoal). Also, th country s strongly dpndnt on ptrolu products portd prarly va th ports of Ivory Coast and Sngal. Ths ptrolu products account for 16% of nrgy consupton. Th avrag rats of annual ncras n th hydrocarbon ports durng th dcad wr approxatly +33% for supr gasoln, +4% for rgular gas and +13% for gas ol. Accordng to stats, th natonal dand for hydrocarbons wll rach on llon tons n 2010, narly 2 llon n 2015 and or than 3 llon n As for lctrcty, ts shar s 3% n th natonal nrgy balanc. In 2004 only 13% of th Malan populaton had accss to lctrcty, wth strong dsparts btwn urban and rural aras. Morovr, ths wak accss s copoundd by a wak shar of rnwabl nrgs. Howvr, Mal has strong potntal n ths fld wth solar nrgy, wnd nrgy and bonrgy (pourghèr, 5 cotton sts, tc). It s only n 2006 that Mal workd out ts natonal stratgy for dvlopng bofuls ovr th prod Th craton of th Natonal Agncy for th Dvlopnt of Bofuls (ANADEB) 6 n 2009 was an ssntal tool for ths stratgy. Anothr rason why th govrnnt has bn actv n prootng ths vgtal product ths last yars s that ths fdstock dos not drctly copt wth food supply. Although larg-scal jatropha plantatons look attractv fro a projct prspctv, thr ndrct pacts on th conoy and th nvronnt ar unknown. Th conoc and 5 Pourghr s th Frnch word for Jatropha curcas L. In Englsh, ths plant s calld physc nut or purgng nut. In Babara (Mal), th usual tr s bagan. In ths artcl, w usd jatropha, bagan and pourghèr, wch ar th coon nas n Mal. 6 ANADEB was cratd n March 2009 wth th andat to: ) dfn tchncal and qualty nors for bofuls, and ontor and valuat ths nors, ) ntat rsarch and dvlopnt of bofuls, ) tran and support prvat oprators n th fld and v) prov nsttutonal fraworks for bofuls and nsur natonal and ntrnatonal partnrshps.

3 nvronntal ssus assocatd wth jatropha, such as food scurty, nrgy dpndncy and th fght aganst sol roson, could b sgnfcant. In trs of food scurty, th an concrn s to antan cral producton, whch s th an stapl of rural Malan housholds (Républqu du Mal, 2008). Thr ar ajor concrns that land and labor oblzaton n th boful sctors would b takn away fro th cral producton sctors, wth consdrabl pacts on food. Ths concrns wr n fact addrssd n a Jun 2008 consultaton, ladng to a rflcton on th rsks and opportunts of th pourghèr ndustry n th country and rprsntd by th MalFolk Cntr, 7 Mal Bocarburant SA, 8 Jatropha Mal Intatv (JMI) 9 and th Group Enrgs Rnouvlabls, Envronnnt t Soldartés (GERES) 10 (Pallèr t al., 2009). Thus, t s crucal to undrstand th pacts of bofuls, such as how thy affct food prcs n th long run. How would such a scalng up of bofuls affct th producton of othr tradabl as wll as non-tradabl coodts? How would t altr th trs of trad? Would larg-scal dvrson of land for boful producton hlp rduc povrty and nqualty n Mal? Th ovrall objctv of ths study s to assss th conoc dvlopnt and nvronntal qualty consquncs of larg-scal xpanson of bofuls n Mal, on of th ky producrs of bofuls n Sub-Saharan Afrca. Ths cas study wll hlp xplan th potntal gans and losss fro th larg-scal xpanson of bofuls at th natonal lvl, takng nto account th opportunty costs of rsourcs and th ovrall pact on conoc prforanc. Th conoy-wd pacts of th larg-scal xpanson of bofuls ar odld usng a CGE odl. Th study asssss th ovrall pact of boful producton on GDP, trad prforanc, prcs, wlfar and conoc growth. Bsds th pacts of xpanson of boful producton to t dostc dand, ths artcl also analyzs varous scnaros rlatd to boful producton, consupton and trad. Th dstrbutonal pacts of bofuls ar odld usng a cro-sulaton odl. Thy ar asurd wth a wd rang of ndcs such as dcoposabl povrty ndcs and nqualty ndcs. Pro-poor analyss ar also prsntd to vrfy whthr th sulatd scnaros bnft or pnalz th poorst Malan housholds. Th Malan odl ncluds 18 producton sctors aong whch sx ar agrcultural sctors, and 4,494 housholds fro th houshold survy ar ncludd n th cro-sulaton odl. Th approach wll allow for all fors of houshold dcoposton. In our odl, w hav takn nto account th rsdntal ara of th houshold had (rural 7 A Malan NGO followng th ntatv of a counal ndustry producng jatropha ol for rural lctrcal dstrbuton (Garalo Coun). 8 A prvat busnss producng bodsl for th Malan transportaton sctor n partnrshp wth th Unon Local ds Socétés Coopératvs ds Producturs d Poughèr (Koulkoro Crcl). 9 A prvat busnss producng pur vgtabl ol for th Malan ndustral arkt, n partnrshp wth th farrs of th Kta Crcl. 10 A Frnch NGO supportng a vllag ndustry producng ol n vw of rural lctrcal dstrbuton (coun of Yorosso).

4 vrsus urban) and dstngushd th rgons n whch jatropha s producd. Fnally, w hav sparatd far and non-far housholds for our analyss. In th nxt scton, Mal s poston n trs of bofuls s prsntd to provd so prlnary ndcatons of rsults fro th analyss prsntd n th lattr part of th papr. Nxt, th an faturs of th cro-sulaton coputabl gnral qulbru odl for th country ar xpland (scton 3). In th fourth part, w dscrb th sulatons prford, followng wth th acroconoc and sctoral rsults (scton 5), whl scton 6 s ddcatd to th analyss of th dstrbuton pacts of polcs on Malan housholds. W conclud n th fnal scton. 2 Bofuls n Mal: an ovrvw Snc a fw dcads, Mal has bn workng on ans to dvlop nw altrnatv nrgy sourcs (alcohol boful and pourghèr, tc.) by ans of varous projcts and progras. Rcntly, Mal undrtook a rfor of ts nrgy sctor by sttng up a natonal stratgy for th dvlopnt of bofuls. Ths stratgy rsultd n th stablshnt of a progra for valung pourghèr cultvatd aras ( ), 11 th laboraton of th Natonal nrgy polcy 12 (2005) and th craton of th Natonal agncy for th dvlopnt of bofuls (ANADEB) (2009). Th jatropha ndustry s stll n ts bgnnngs, howvr. Accordng to GEXSI (2008), vry fw projcts n 2008 xcdd two yars n lngth and nvolvd sgnfcant ol producton. Although thr dos not yt xst a arkt for th foral corcal xchang of jatropha and ts ol, any countrs ncludng Mal ar bcong awar of th potntal of ths plant n trs of producng grn nrgy and substtutng for hydrocarbons. Jatropha: proprts and uss Th pourghèr plant blongs to th Euphorbaca faly, whos sds ar blvd to hav bn portd by th Portugus fro Cntral and South Arca n th 17th cntury. Th shrub s frquntly cultvatd n hdgs n ordr to fght aganst th watr and wnd roson of sols and to protct agrcultural crops fro rovng anals (Hnnngs, 2009). It was cultvatd n th for of quckst hdgs usd as nclosurs for cral and vgtabl gardns, partcularly around vllags. Anothr rason xplanng th ntrst n pourghèr s ts rsstanc to drought, whch s frqunt n Mal. Indd, ths plant now grows n ost of th plant s s-dsrt rgons that rcv btwn annual ranfall and 11 Th Progra Natonal d Valorsaton Énrgétqu d la Plant Pourghèr (PNVEP) cos wthn th scop of Malan polcs and stratgs prsntd n th Progra Natonal d Acton Envronnntal (PNAE), th Cadr Stratégqu d Lutt contr la Pauvrté (CSLP) and th prooton of th Stratég Natonal ds Énrgs Rnouvlabls. It s plntd by th Cntr natonal d l énrg solar t ds énrgs rnouvlabls (CNESOLER) undr th suprvson of th Drcton Natonal d l Énrg of th nstèr d l Énrg, ds Mns, t d l Eau. 12 Th Natonal nrgy polcy as to ncras local producton of nrgy by dvlopng bofuls to satsfy th sococonoc nds of th country at a lowr cost.

5 xprnc long drought prods. It s cultvatd usng drct drllng of sds, prckng out nto pots (plant nursrs) or cuttag (Latap, 2007). Ths plant has also bn xplotd snc svral dcads to obtan olagnous sds contanng btwn 20% and 40 % ol dpndng on th xtracton thod usd (anual, chancal or chcal) (Lgndr, 2008). Ths ol can b usd for lghtng and cookng, as wll as for producng soap, bocds, sd caks or ful (S Fgur 4). It bcos a sutabl dsl substtut for us n a wd rang of statonary ngns usd n rural aras, such as rrgaton pups, lls and gnrators (Achtn t al., 2008). Jatropha around th world and n Mal Th cultvaton of pourghèr was ntroducd n Mal by Frnch colonalsts at th nd of th 1930s n th Offc du Ngr and popularzd n 1990s by th Gran dvlopnt coopraton (GTZ). Th plant ca to b usd as a barrr protctng agrcultural aras fro lvstock (shp, goats, tc.), sol roson du to watr run-off and wnd, partcularly n rural aras. It was also usd by won as tradtonal dcn aganst alara or gnral pan. On a global scal n 2008, approxatly 900,000 ha wr countd, of whch 85% wr apparntly plantd n Asa (Indonsa, Myanar, Inda, Chna). Th shrub was thought to covr 120,000 ha n Afrca, prarly n Madagascar, Zaba, Tanzana and Mozabqu. By 2015, world producton could approach 13 llon hctars. In trs of Afrca, Ghana and Madagascar alon ay count 1.1 llon of th two llon hctars xpctd ovr th ntr contnnt (GEXSI, 2008). 13 It s also ntrstng to not that far szs ar on th ncras. In 2008, only 11% of jatropha producton n Afrca was accountd for by larg plantatons. GEXSI (2008) prdcts that n 2013, 38% of producton on th contnnt wll b accountd for by larg fars, ncludng 23% n plantatons of or than 1000 hctars (Fgur 1). Fgur 1: Dvlopnt of Jatropha Schs by 2013 Sourc: (GEXSI, 2008). 13 Ths fgurs wr statd by th Global Exchang for Socal Invstnt (GEXSI) sponsord by th World Wldlf Fund (WWF) to carry out a global arkt study on jatropha. Th goal was to prsnt th stuaton of jatropha cultvaton n 2008 and to fors th dvlopnt of projcts fro now to 2015 for Latn Arca, Afrca and Asa. (Cf.

6 In 2002, Mal had at ts dsposal roughly 10,000 klotrs of jatropha shrubs, wth a growth rat statd at approxatly 2,000 klotrs pr yar (Latap, 2007). In 2008, consultd xprts countd 1,800 hctars of pourghèr shrubs, wth an objctv on th part of producrs of rachng 23,000 hctars by 2015 (GEXSI, 2008). Howvr, as ntond n th agrcultural rport of th OECD and FAO (OCDE-FAO, 2009), Afrcan projcts td to th xpanson of th ndustry ar stll odst n Afrca. In 2009, Pallèr t al. statd that th total cultvatd surfac rachd 3,730 ha (Tabl 1) dstrbutd n four sgnfcant producton aras. Th frst plac occupd x aquo by Kta and Koulkoro (34.5%) followd th hstory of ths rgons n th cultvaton of quckst hdgs of pourghèr. Fnally, th Yorosso and Garodo rgons rprsnt 31% of cultvatd surfacs n th country. Tabl 1: Surfac of cultvatd pourghèr n Mal Aras Surfac (ha) % # of producrs % Kta 1 300,00 34,5% 1 313,00 42% Koulkoro 1 300,00 34,5% 1 017,00 32% Garalo 430,00 13% 530,00 17% Yorosso 700,00 18% 300,00 9% Total 3 730,00 100% 3 160,00 100% Sourc: Pallèr t al. (2009) In Mal, th qust for transforng gran nto boful s not a nw concrn, as t startd n th bgnnng of th 20 th cntury n th rgon of Sgou (Haïdara, 1996). Th proprts of ths boful ar vry clos to thos of so dsl fuls. Furthror, th rturn of ths plant s ntrstng, as t can rach 1,900 ltrs of ol pr hctar vrsus 572 ltrs/ha wth colza, 662 ltrs/ha wth sunflowr or 446 ltrs/ha wth soya (Lgndr, 2008). Snc pourghèr producton taks plac on a sall scal n Mal, th four prvously ctd aras appar to stand out clarly (Pallèr t al., 2009). 14 In 2006, th Mal FolkCntr dvlopd a projct targtng th producton of jatropha ol n vw of rural lctrcal dstrbuton (gnratng sts) n th agrcultural and cotton rgon of Garolo. Bsds cotton, ths rgon s also known for th producton of panuts and portant food crops (corn, sorghhu, rc, tc.). In ts projct launchd n 2007, Mal bocarburant SA usd th proprts of jatropha to produc bodsl ntndd for th dostc arkt n th Koulkoro Crcl. Ths rgon, whch strtchs ovr 7,000 k 2, cobns th producton of corcal crops (cotton, panut, and tabacco), food crops (corn, llt, rc) and truck farng thanks to ts proxty to th Ngr Rvr. In th sa yar, th Jatropha Mal Intatv (JMI) projct was also launchd n Kta n th rgon of Kay. Th projct ntally targtd th producton of bodsl. Snc 2009, jatropha vgtabl ol dstnd for th Malan ndustral arkt has bco th frst output rsultng fro ths sd transforaton. Fnally, n 2008, th Group Énrgs Rnouvlabls, Envronnnt t Soldartés (GERES) plntd a jatropha producton projct n th Koutala Crcl for rural 14 Cf. Fgur 5 for a gographcal ovrvw of ths jatropha producton aras.

7 lctrcal dstrbuton purposs. 15 Although, sn fro ths angl, jatropha producton appars to b an ntrstng avnu for crtan populatons, th ltratur rvals an absnc of consnsus n trs of ts concrt conoc bnfts. Th quston s whthr xpandng ths typ of producton should b advocatd or, on th contrary, th sctor should b dvlopd prudntly. Jatropha dvlopnt: th controvrsy Aong th objctvs ost oftn assocatd wth projcts for dvlopng jatropha cultvaton, th advancnt of won (producton of soap and sal of ol), th rducton of vllag povrty (crop protcton, sal of sds, ol, soap), sol roson control (plantng of lv shrubs) and th producton of rnwabl nrgy ar ost oftn ctd (Hnnng, 2009). Fnally, pact analyss of ntroducng and dvlopng ths cultvaton n rural sttngs on th wlfar of housholds and counts s also bgnnng to gnrat growng ntrst n th currnt contxt of Mllnu Dvlopnt Goals (Pallèr t al., 2009). Crtan rsarchrs such as Matthws (2007) s th producton of bodsl fro jatropha as a ans to ndustralz poor countrs and hlp th ovrco povrty. Plantd on a larg scal, jatropha ay vn nabl ths countrs to achv nrgy autonoy, as wll as ncras thr producton and xportaton. Rnnr (2007) wll suarzs th prcpton of any as, accordng to hr, Jatropha ss to offr th bnfts of bofuls wthout th ptfalls. Howvr, othrs quston th producton of bofuls fro agrcultural products. Thy ras concrns ovr bodvrsty loss (Razon, 2009) or ndangrng th food sovrgnty of countrs that ar alrady strugglng (Luoa, 2009; Razon, 2009; Capa and Valntn, 2009). Othr concrns nclud th possbl nflunc of pourghèr plantatons on sol frtlty or th rsks of propagaton of dsass or parasts that could annhlat not only jatropha but also othr narby culturs (Capa t al., 2009). 16 Ths ltratur shows that two clans ar at odds whn t cos to jatropha producton. Nvrthlss, as Bng (2006) stats, thr s no ndcaton that any scntfc studs hav bn conductd to prov or dsprov any of th abov. Basd on th fact that th Afrcan contnnt stll dsposs of a sgnfcant volu of unxplotd agrcultural land (approxatly 58% accordng to GEXSI, 2008), t appars rlvant n our vw to dvlop ths rar analyzs, partcularly n trs of xanng wlfar and food scurty. Ths s what w propos n th followng sctons through an analyss of th cas of Mal. 15 In our analyss, w wll stablsh th xstnc of a jatropha producton ara n Mal basd on ths four projcts. 16 Th authors would lk to rcall th Afrcan coff plant (arabca) ntroducd to th ntr ntrtropcal ara by colonsts and whos plantatons wr dcatd n th arly 20th cntury by orang rust.

8 3 Th Malan odl 3.1 Th data Th SAM was constructd wth th an data drawn fro an nput output tabl wth 18 productv sctors. 17 Th rfrnc yar for th SAM s W xtractd th jatropha sctor fro th cash crop sctor and th ful sctor was dcoposd nto fossl fuls and bofuls. W had to dsaggrgat th n ordr to focus our analyss on ths sctors. Scondary data was usd fro varous sourcs to dcopos ths sctors. Th nd rsult was a SAM wth 20 producton sctors. Th nxt stp was to lnk th cro-houshold data 4,494 housholds fro th 2006 Enquêt légèr ntégré auprès ds énags (ELIM-2006) survy to th SAM. W odfd th nco and xpndtur structurs for th housholds basd on th nonclatur of th SAM. 3.2 Th odl Snc th lat 1990s rsarchrs hav bn usng CGE to analyz th pact of polcy rfors on povrty and nco dstrbuton. Thr an catgors of ths odls hav bn usd durng ths prod: th rprsntatv houshold approach (RH), th ntgratd ult-houshold approach (IMH) and th ult-houshold squntal approach (MHS). Th CGE-RH approach dvds housholds nto groups, choosng a rprsntatv houshold for ach group and usng that rprsntatv houshold n th CGE odl. Th varatons n nco of th rprsntatv agnts gnratd wth th CGE odl ar appld to housholds wthn thr rspctv group fro a houshold survy. Ths assupton dos not allow th analyst to tak nto account wthn-group changs n nco dstrbuton, vn though studs (Hupp and Ravallon 1991 and Savard 2005, for xapl) hav shown that such changs can b gratr than btwn-group nqualty changs. Ths s tru both for th statc asur and for varatons followng polcy sulatons. Savard (2005) donstratd that th rsults of povrty and nco dstrbuton analyss can b copltly rvrsd by takng nto account wthn-group dstrbutonal ffcts. To solv ths probl, a scond approach was proposd by Dcaluwé, Duont and Savard (1999) and appld by Cognau and Robllard (2000), Gørtz t al. (2000) and Cockburn (2001), naly th CGE ntgratd ult-houshold approach (CGE-IMH). Ths thod ncorporats a larg nubr of housholds fro a houshold survy (and sots all of th) nto th CGE odl. Th approach taks nto account wthn-group dstrbutonal ffcts and has th furthr advantag of provdng cohrnc btwn th cro and acro parts of th odl, but at a cost. Frst, data rconclaton can b vry problatc (Ruthrford, Tarr and Shpotylo, 2005); scond, nurcal rsoluton can b challngng (Chn and Ravallon, 2004). 17 Th orgnal nput-output atrx was obtand fro th Drcton natonal d la statstqu t d l'nforatqu.

9 Th thrd approach s rfrrd to as th CGE cro-sulaton squntal thod (MSS) and could b subdvdd nto two varants. Th frst, cro-accountng, was forally prsntd by Chn and Ravallon (2004) and has bn xtnsvly appld n rcnt yars. 18 Th scond varant, proposd by Bourgugnon, Robllard and Robnson (2005), conssts of ntgratng, at an ndvdual lvl, rch cro bhavor obsrvd at a houshold lvl such as consupton or labor supply. Ths vrson ntroducs or htrognty btwn housholds wth th applcaton of a croconotrc odl. Th gnral da of th MSS approach s that a CGE odul fds arkt and factor prc changs nto a crosulaton houshold odl. Th an crtcs lvld at ths approach s that th crofdback ffct s not fully takn nto account; th quston has bn rasd n two ltratur rvws of acro-cro odlng for povrty analyss (Hrtl and Rr (2005) as wll as Bourgugnon and Spadaro (2005)). Howvr, Bourgugnon and Savard (2008) found that th loss of nforaton assocatd wth usng th MSS approach can b rlatvly sall and polcy conclusons wr robust btwn th two approachs. 19 In ths papr w apply th cro-accountng vrson of th MSS approach. 20 Th last two approachs (IMH, MSS) allow for rch analyss of nco dstrbuton and povrty bcaus thy nclud a larg nubr of housholds n th odlng xrcs. Ths n turn allows th odlr to apply povrty and nco dstrbuton asurs and ndxs followng polcy sulatons. As alrady ntond, th IMH approach s th soundst on a thortcal bass. Howvr, wth ths approach, t s ncssary to construct a balancd subatrx for houshold accounts n a standard socal accountng atrx. On th othr hand, wth th CGE-MSS approach, th houshold nco and xpndtur do not rqur balancng. Ths gan n flxblty cos fro th fact that th cro odul s solvd squntally and that w us prc prcntag changs to lnk th CGE odul to th cro houshold odul. Ths constrant has ld us to slct th CGE-MSS approach for our analyss. W wll rfr to th houshold sub atrx databas as th houshold odul. To captur th pact of jatropha and boful polcs on th wlfar of ndvdual housholds and dstrbutonal ffcts, w had to adapt a standard odl to spcfcts of ths two sctors n th Malan conoy. Th prsntaton of th odl s dcoposd nto th CGE odul and houshold cro houshold odul CGE odul Th CGE coponnt of our odl (hraftr rfrrd to as th CGE odul) s basd on th EXTER odl of Dcaluwé t al. (2001), wth sgnfcant adjustnts, whch w wll dscrb aftr prsntng th an faturs of EXTER. Th producton structur s strongly 18 Aong arly applcatons of ths approach ar Vos and D Jong (2003) and Kng and Hnda (2003). 19 Bourgugnon and Savard s (2008) coparatv analyss of th IMH and MSS approachs was appld to th Flpno conoy. In ths study, th labor supply was ndognous and th largst porton of th gap n th rsults obtand fro th two approachs ca fro th labor supply. In our applcaton, th labor supply wll b hld constant. 20 Th an rason for slctng ths approach s th lack of adquat data to apply th croconotrc approach.

10 odfd. 21 W prsnt th ultlayrd producton structur n th followng graph to splfy th dscrpton. Fgur 2: Th CGE odl and ts structur Startng at th top lvl, w hav total producton of th sctor (XS), whch s ad up of fxd valu-addd shars (VA) and ntrdat consupton (CII), as gnrally assud n standard CGE odlng. Th rlatonshp dtrnng th lvl of VA s a Cobb-Douglas producton functon btwn copost labor (LD) and total captal (KDT). Producrs nz thr cost of producng VA subjct to th Cobb-Douglas functon. Frst ordr condtons fro ths procss ar usd to dtrn th optal labor dand quatons. Labor (LD) s thn dcoposd nto sklld (LDQ) and unsklld (LDNQ) labor, wth th cobnaton of ths two factors bng dtrnd by a constant lastcty of substtuton (CES) functon, onc agan through a procss of cost nzaton. Ths assupton pls that changs n th rlatv wags of th two typs of labor wll lad th producr to odfy th rato btwn th two groups of workrs, subjct to constrants on substtuton lnkd to hs producton capacts. Th total captal (KDT) s dcoposd nto land (LAND) and captal (KD). Th sa CES functon s usd to lnk ths two factors. Th captal s sctor-spcfc (fxd) and th land can b thr obl btwn agrcultural sctors or fxd. 22 Movng on to ntrdat consuptons, w sgnfcantly nrch th EXTER odl by dcoposng ths nputs nto varous sub-nputs. W frst hav th total 21 W provd th coplt st of quatons n th appndx. 22 W wll us th two assuptons n our sulatons.

11 ntrdat consupton (CII) that s dcoposd nto nrgy (CIE) and othr ntrdat consuptons (CI). Ths two sub-nputs ar lnkd wth a CES functon. Th CIE s furthr dcoposd nto fuls (FUEL) and othr nrgs (ENER). Ths two nputs ar lnkd wth a CES functon, thrfor allowng for substtuton n ths nputs. Th fuls ar furthr dcoposd nto fossl fuls (FFUEL) and bofuls (BFUEL) wth th sa functonal for (CES). Othr ntrdat consupton (CI) was odld as fxd shars fro th nput/output ratos calculatd fro th data n th SAM. Ths s a coon practc wth ths typ of odl. Th rch structur wll allow bofuls to ncras thr arkt shar to rplac two typs of nrgs as an ntrdat nput n th producton procsss of othr sctors of th conoy. Sctor-spcfc lastcts of substtuton ar usd to rflct dffrncs aong sctors n dtrnng th x of factors n ach CES functon. W usd hghr lastcts for substtuton btwn fossl fuls and bofuls copard to fuls and othr nrgs. W allowd for ths substtuton as our rvw rvals varous potntal uss of bofuls as a sourc of nrgy. As notd arlr, our odl for Mal covrs 20 producton sctors, ncludng jatropha, bofuls, fossl fuls and othr nrgs (anly lctrcty). Jatropha and bofuls wr not found n th ntal SAM (Coulbaly 2009). Th sz of ths sctors s rlatvly sall n Mal but thy prsnt substantal potntal for growth. In th odl, four agnts ar odld. Frst, th govrnnt draws ts rvnus fro producton taxs, port duts, housholds and prvat frs nco taxs as wll as fro transfrs fro th rst of th world (budgtary assstanc). Its xpndtur s ad up of th consupton of publc srvcs and of transfrs to othr agnts. W hav on agnt rprsntng aggrgat prvat frs, whch draw thr nco fro captal nco and pay taxs; thy ak transfrs to othr agnts and sav for nvstnt. Th rst of th world s consdrd as anothr agnt n a standard fashon as t s usd to odl conoc rlatons btwn Mal and th rst of th world ncludng ports, xports and transfrs to and fro agnts n Mal. Fnally, w nclud a sngl rprsntatv houshold n th CGE odul. Ours s a odl of a sall opn conoy to whch world prcs of ports and xports ar xognous. W posd th Arngton hypothss (1969) for port dand, whrby dostc consurs can substtut dostcally producd goods wth ports (prfctly) accordng to an lastcty of substtuton that s sctor spcfc. Whr local consurs hav no prfrnc btwn portd and local goods, w us a hgh lastcty of substtuton; convrsly, th lastcty of substtuton s low whr consurs prfr on good ovr anothr. 23 Th rlatv prc of th two goods s th othr dtrnant of th rato of dand for portd goods vrsus dand for local goods. On th xport sd, producrs can sll th goods on th local arkt or xport thr producton and ar nfluncd by 23 Elastcts of substtuton wr slctd aftr a consultaton procss wth cvl srvants fro th nstry of fnancs, statstcs offc and trad nstry.

12 rlatv prcs on ach arkt and by thr rspctv lastcty of transforaton of th good for on or th othr arkt. Our prc quatons ar standard othr than for utlts. W usd th nonal xchang rat as a nurar, and, as statd arlr, ntrnatonal prcs (ports and xports) wr xognous. Accordngly, th country has no control ovr th prcs appld on th world arkt. Our odl s qulbru condtons ar also standard. Th coodty arkt s balancd by an adjustnt of th arkt prc of ach coodty. Th labor arkt s prfctly sgntd and balancs out wth an adjustnt of th nonal wag on ach of th rspctv arkts (sklld and unsklld). It s thrfor possbl for workrs to ov fro on sctor to anothr, but not fro on arkt to anothr. Labor supply n ach of th arkts s fxd, and thr s no unploynt. 24 Th currnt account balanc s xognous, and s balancd wth an adjustnt of th ndognous prc ndx (GDP dflator). Wth rgard to th qulbru of savngs and nvstnt, total nvstnt adjusts to th su of th savngs of all agnts n th odl Mcro Houshold odul In th cro houshold odul, w nclud all of th 4,494 housholds fro th survy (EMEP-II). W hav spcfd nco and xpndtur functons for th housholds, whch ar paratrzd on th houshold-spcfc nforaton found n th survy. As ntond, th odul s solvd squntally (CGE odul and houshold odul). Lt us now dscrb ths squnc. W frst spcfy an nco quaton that rflcts nco structur for ach houshold n th EMEP--II survy. W assu that th ndownts of factors ar xognous. W us th factor paynt varatons gnratd by th CGE odul and apply th to th factor ndownts. Th two wag rats ar also appld to labor ndownts of housholds. 25 As n th CGE odul, w assu that th transfrs fro othr agnts to th housholds ar xognous. Ths procdur provds us wth th nw houshold-spcfc nco. W can thn ov on to th xpndtur sd. Th portant coponnt for wlfar analyss s th chang n houshold consupton. Th dand functons ar drvd fro a utlty axzaton procss (Cobb-Douglas utlty functon), and ths dand quaton s a functon of arkt prcs and houshold nco. Th fnal stp n th squntal rsoluton of th houshold odul conssts n coputng th chang n wlfar. Iplctly, ths allows us to tak nto account sultanously th nco and prc ffcts on ach houshold s wlfar. 26 As n Bourgugnon and Savard (2008), w us th varaton of th ral nco to asur th 24 Ths dos not an that w assu zro unploynt n th conoy, but rathr that unploynt s xognous to th odl. 25 Th survy ncludd nforaton that allowd us to dcopos th labor nto qualfd and non qualfd labor. 26 Ths approach s dffrnt fro th ndognous povrty ln proposd by Dcaluwé t al. (2005) as t capturs th prc ffct of th sulaton through spcfc houshold prfrnc and not through a basc nds approach. For a dscusson of th advantags and dsadvantags of th two approachs, s Ravallon (1998).

13 chang n wlfar. Th houshold-spcfc valu shars for consupton ar coputd fro obsrvd fgurs n th survy. Ths shars ar thn usd to spcfy a houshold-spcfc prc ndx that s n turn usd wth th nonal nco chang to coput wlfar changs Povrty and dstrbutonal analyss Aftr dntfyng th targt 27 groups for th bas prod and sulatons, th nxt stp s to calculat and copar povrty and nqualty ndcs. Th ndcs chosn for povrty and nqualty ar d Fostr, Grr and Thorbck 28 (F-G-T, 1984) P α and Gn, rspctvly. In addton to ths ndcs, w appld pro-poor growth analyss Sulatons W prfor fv an sulatons wth thr varants of th frst thr sulatons. Our frst thr sulatons consst n ncrasng producton of jatropha by provdng nw land for ths purpos. W prford a fftnfold ncras n land for jatropha producton gong fro 3,000 to 45,000 hctars. Ths s a larg ncras but t only rprsnts 1% of th othr agrcultural land n Mal out of approxatly 4 llon hctars of xplotd arabl land. Thr optons ar sulatd. In th frst opton, th xpanson of land s accoplshd by usng dl land and hnc puttng no prssur on othr fors of agrcultur. In th scond scnaro, w assu that th ncras n producton s gnratd by usng half dl land and half fro othr agrcultural sctors. In th thrd sulaton, w assu that all th land rqurd for th xpanson of th jatropha sctor wll b takn fro othr agrcultural sctors. Ths wll provd us wth uppr and lowr bounds for th ncras n land us for jatropha producton. In th nxt sulaton (sulaton 4a), w provd a subsdy of 20% to th boful producton sctor and n th last sulaton, 4b, w cobn ths subsdy wth sulaton 1a. Sulatons 1b, 2b and 3b ar dffrnt than odls 1a, 2a and 3a at th cro houshold odl lvl. In th frst scnaros, w assu that th xpanson wll b prford unforly by farrs n th jatropha ara ths ar scnaros 1a, 2a and 3a and n th scond, th farrs of all rgons wll produc or jatropha or start producng jatropha. Tabl 2 suarzs ths sulatons. 27 Ths groups hav bn dntfd n an ffort to rflct th groups prsnt n and outsd of th Jatropha ara, to assss pact on ths groups and to copar thr pacts on th othr groups n th Malan conoy. It should b notd that th dffrnt groups could vry asly b ntroducd n our analyss. 28 FGT povrty ndcs ar addtvly dcoposabl; as such, thy ar ntrstng wthn th frawork of ths analyss and ak t possbl to asur th proporton of th poor aong th populaton but also of ths povrty dpth and svrty. Th hghr th dgr of povrty avrson, th gratr th portanc grantd to th poorst. For dtald nforaton on ths ndx faly, rad Ravallon (1994). 29 A dtald vrson of ths pact analyss wll b basd on thr ndcs: th pro-poor growth ndx (PPGI) of Kakwan and Prna (2001), th povrty quvalnt growth rat (PEGR) of Kakwan and Son (2002), and th absolut rat of pro-poor growth of Ravallon and Chn (2003) drvd fro th GIC. For a dtald prsntaton of ths ndcs, s Boccanfuso and Ménard, 2008.

14 Cod 1a 1b 2a 2b 3a 3b 4a 4b Tabl 2: Sulaton dscrpton Sulaton dscrpton 15 ts ncras n agrcultural land usng dl land n jatropha ara 15 ts ncras n agrcultural land usng dl land n Mal 15 ts ncras n agrcultural land usng 50% dl land n jatropha ara and 50% agrcultur land 15 ts ncras n agrcultural land usng 50% dl land n Mal and 50% agrcultur land 15 ts ncras n agrcultural land usng 100% agrcultur land 15 ts ncras n agrcultural land usng 100% agrcultur land 20% subsdy fundng by dcras n govrnnt xpndtur 20% subsdy fundng by dcras n govrnnt xpndtur & sulaton 1a 5 Rsults W procd wth a brf dscrpton of our acro and a fw sctoral rsults. As w focus our analyss on a dstrbutonal pact of ths scnaros, w slct so ky varabls that play an portant rol n ths dstrbutonal analyss. 5.1 Macroconoc and sctoral pacts Fro our rsults prsntd n Tabl 3 blow, th frst gnral statnt w can forulat s that th pact on acro varabls s rlatvly sall. Ths s not surprsng as th two an sctors of ntrst ar rlatvly sall at th rfrnc prod. Th frst sulaton s th on wth th wakst pact at th acro lvl. Hnc, usng dl land wll hav a slght postv pact on all acro varabls. Whn th jatropha sctor copts wth othr agrcultural sctors for land, w obsrv a rducton n all acro varabls. Ths ffcts ar aplfd wth usng all th land for xpanson fro othr sctors. W also rtrat that th captal rqurd by th jatropha sctor wll b drawn fro othr agrcultural sctors. Hnc, th coptton cos fro th land and captal sd for othr agrcultural sctors. It s also portant to nton that w assu fxd labor supply and hnc, th xpanson of on sctor cos at th xpns of th othrs. Th fourth sulaton (4A) s ntrstng as t provds postv rsults for ost varabls. Th only ngatv pact s for qualfd wag, whch s th rsult of layoffs n th publc sctor to fund th subsdy. Othr varabls ncras wth th othr xcpton bng th provnt of th currnt account dfct (a rducton n th dfct). W fnd th strongst postv pact on total nvstnt, non-qualfd wag and rntal rat of agrcultur. It s ntrstng to not th opposng pact on qualfd wag that dcrass and non-qualfd wag that ncrass. Ths spcfc ffct should bnft th poorr housholds vrsus th rchr ons as th lattr ar or prvalnt n th qualfd arkt. Th last sulaton s qut slar to sulaton 4A but wth a slghtly or postv pact on all varabls xcpt th currnt account balanc and th non-qualfd wag.

15 Tabl 3: Macro lvl rsults of th CGE odl Varabls Dfnton Rfrnc S 1A S 2A S 3A Sulaton 4 (A & B) Boful Subsdy Boful subsdy & 1A Houshold Yh nco ,01-0,20-0,42 0,04 0,05 w 1 qualfd wag 1 0,03-0,05-0,12-0,70-0,62 w 2 non qualfd wag 0,5 0,00-0,29-0,57 0,15 0,14 rntal rat of rr agrcultural captal 1 0,03-0,69-1,41 0,09 0,11 Govrnnt Yg nco ,01-0,15-0,31 0,00 0,00 Govrnnt Sg savngs ,02-0,56-1,15 1,50 1,42 currnt account cab balanc ,16-1,23-2,67-2,62-2,21 Y Frs' nco ,02-0,19-0,40 0,07 0,09 S Frs' savngs ,03-0,39-0,82 0,15 0,18 It total nvstnt ,06-0,74-1,56 0,64 0,67 Gross dostc product ,02-0,22-0,45 0,00 0,02 GDP Movng on to th sctoral rsults, onc agan w hav vry ltd pact wth th xcpton of th jatropha and boful sctors. Ths s th rsult of th 15 fold ncras n land us n th lattr sctor that producs or nput for th boful sctor at lowr cost. Lookng at th output/valu addd changs (n Tabl 4) w not that for th frst sulatons, th sctors dcrasng output ar th fossl fuls, and to a lssr xtnt th constructon, publc srvcs and srvc sctors. Othr sctors vry slghtly ncras thr producton. In th scond and thrd sulatons, drawng land fro othr sctors has a strongr pact and ths ar ngatv for ost sctors. Howvr, th nng and anufacturng sctors as wll as th publc srvcs ncras thr producton. Th thrd sulaton producs strongr rsults. In ths two sulatons, th two sctors ost ngatvly affctd ar th cotton and crop agrcultur sctor, and th lvstock sctor, th last ngatvly affctd. For th last two sulatons, w not that th subsdy s farly ffcnt n gnratng an ncras n output for th jatropha sctor and boful sctors, whch wr both ncrasd by or than 10%. Th publc sctor s th on dcrasng th ost as ths subsdy s fundd by a rducton n publc srvc provson. Th prc ffcts follow a slar trnd wth wak ffcts n gnral for all sctors but jatropha and bofuls, and wakst pact for sulaton 1 and th strongst ffcts for sulaton 4B.

16 Tabl 4: Macro and sctoral lvl rsults of th CGE odl Varabls Dfnton Rfrnc Va (valu addd) Sulaton 4 (A & B) S S S 1A 2A 3A Boful Boful Subsdy subsdy & 1A crop agrcultur ,00-0,76-1,52 0,00 0,00 rc ,02-0,36-0,74-0,02 0,00 xport ,02-0,59-1,19 0,01 0,03 jatropha ,06 89,79 89,51 10,37 110,23 coton ,02-0,75-1,51 0,17 0,17 Lvstock ,02-0,24-0,50 0,02 0,04 forstry ,00-0,61-1,21-0,02-0,02 nng ,00 0,02 0,04 0,08 0,07 agondustry ,00 0,01 0,02 0,03 0,03 txtl ,01 0,03 0,05-0,02-0,01 othr ndustrs ,00 0,08 0,15-0,03-0,03 othr nrgy ,00-0,01-0,02 0,01 0,01 boful ,93 84,71 84,48 10,83 105,66 constructon ,02-0,15-0,28 0,11 0,08 corc ,00-0,01-0,03 0,00 0,01 transport ,01 0,01 0,01 0,04 0,05 srvcs ,01-0,02-0,04 0,00-0,01 bankng & fnanc ,00-0,09-0,17 0,09 0,08 publc srvcs ,01 0,03 0,08-0,38-0,38 fossl fuls ,05-0,05-0,05-0,05-0,10 As w can s fro Tabl 5, th strongst pact s on jatropha (-77%); gvn th strong ncras n supply, a drop n prc s ncssary for th conoy to absorb ths xcss supply. Snc ths prc dcrass sgnfcantly and t s th an nput to produc boful, ths prc also dcrass by 39% for th frst thr sulatons. For th fourth sulaton (4A), th prc of bofuls dcrass by clos to 8% and as a rsult of th subsdy, ths ncras n copttvnss allows th sctor to ncras ts output by alost 11%. Ths ncras n producton wll rqur an ncras n dand for jatropha and gvn th fxd land assupton, ths lads to a strong ncras n th prc of jatropha of 15%. Whn w cobn ths sulaton wth th 1A sulaton w gratly rduc th dand sd prssur on jatropha and fnd alost th sa dcras n th prc of jatropha. Th two sulatons cobnd aplfy th prssur on th prc of bofuls wth a dcras just gratr than 43%. It s also ntrstng to not that th arkt prcs of othr agrcultural goods all ncras as a rsult of any on of th sulatons prford. Tabl 5: Macro and sctoral lvl rsults of th CGE odl

17 Varabls Dfnton Rfrnc S 1A Pq (arkt prc) S 2A S 3A Sulaton 4 (A & B) Boful Subsdy Boful subsdy & 1A crop agrcultur ,02 0,50 0,98 0,11 0,12 rc ,01 0,18 0,38 0,02 0,01 xport ,02 0,40 0,78 0,04 0,06 jatropha ,46-77,51-77,55 14,99-76,96 coton ,01 0,11 0,20 0,16 0,16 Lvstock ,01 0,01 0,02 0,05 0,04 forstry ,01 0,36 0,71 0,04 0,06 nng ,01-0,08-0,16-0,02-0,01 agondustry ,00-0,14-0,28-0,06-0,05 txtl ,01-0,23-0,47 0,08 0,08 othr ndustrs ,00-0,05-0,09-0,01-0,01 othr nrgy ,00-0,30-0,60 0,09 0,08 boful ,81-39,03-39,24-7,94-43,63 constructon ,01-0,25-0,50 0,11 0,09 corc ,04-0,39-0,83 0,09 0,13 transport ,01-0,12-0,25-0,01 0,00 srvcs ,01-0,24-0,47 0,02 0,01 bankng & fnanc ,01-0,21-0,41-0,18-0,18 publc srvcs ,02-0,11-0,23-0,41-0,36 fossl fuls ,01-0,21-0,42 0,03 0,02 W coplt ths scton wth a look at th rntal rat of captal for th non agrcultural sctors (Tabl 6). Snc th captal s fxd, ach sctor has a spcfc rntal rat of captal. In th frst sulaton, ost sctors do not xprnc notcabl changs n th rntal rat of captal. Th strongst changs ar for fossl fuls (a copttor of bofuls), th transportaton sctor (an portant consur of ful), corc and constructon sctors. As for othr varabls, sulatons 2A and 3A produc strongr ffcts. In th thrd sulaton w obsrv a fw varatons gratr than 1% n absolut trs wth a dcras of 1.45% n th constructon sctor and a 1.05% dcras n th corc. Th subsdy (sulaton 4A) producs a postv pact on th boful sctor but wth a uch sallr pact copard to th frst thr sulatons. Th last sulaton producs a vry hgh pact on th boful sctor. Such an ncras n captal rturn should gnrat ajor nvstnt n th sctor n th futur but w do not captur ths as w hav a statc odl. Ths ncras n nvstnt would sgnfcantly rduc th rturns n ths sctor. Tabl 6: Macro and sctoral lvl rsults of th CGE odl

18 Varabls Dfnton Rfrnc r (rntal rat of captal) S 1A S 2A S 3A Sulaton 4 (A & B) Boful Subsdy Boful subsdy & 1A nng 1-0,01 0,10 0,20 0,07 0,06 agondustry 1 0,01-0,17-0,35-0,09-0,07 txtl 1 0,02-0,22-0,45 0,06 0,08 othr ndustrs 1 0,00 0,04 0,08-0,13-0,13 othr nrgy 1 0,00-0,31-0,63 0,10 0,10 boful 1 541,88 538,32 534,73 36,13 782,91 constructon 1-0,05-0,75-1,45 0,47 0,39 corc 1 0,06-0,49-1,05 0,13 0,18 transport 1 0,14 0,09 0,05 0,01 0,15 srvcs 1-0,04-0,36-0,68 0,11 0,07 bankng & fnanc 1 0,02-0,30-0,62-0,41-0,36 publc srvcs 1 0,00 0,00 0,00 0,00 0,00 fossl fuls 1-0,15-0,22-0,30-0,80-0,90 W wll follow ths scton wth th dstrbutonal analyss. It wll b portant to kp n nd th varatons prsntd n th prvous tabls. Morovr, as w hav dscrbd n th ntroducton of ths scton, w apply two sub scnaros n th crohoushold n whch w assu n sulatons A that th xpanson of jatropha producton s don n th jatropha ara whras n th sulatons B th ncras s appld proportonally to all far housholds n th survy. 5.2 Dstrbutonal pacts In ths scton, w prsnt and analyz povrty and nco dstrbuton changs followng th dffrnt polcy scnaros. Th data hr usd has bn drawn fro th 2006 Enquêt Légèr Intégré auprès ds énags (ÉLIM) nvolvng 4,494 housholds across th Malan trrtory. W prfor ths analyss at th natonal lvl and dcopos housholds basd on thr crtra. W frst dstngushd btwn far and non-far housholds. To ntroduc gratr htrognty, w thn dstngushd btwn housholds locatd n aras n whch svral jatropha producton projcts hav bn plantd and othrs. 30 Fnally, w dstngushd btwn rural and urban housholds. In both th crosulaton odl and th wlfar analyss, w hav thus carrd out our work basd on ght houshold groups: 1) Far housholds lvng n th urban ara whr jatropha s producd (MAJU); 2) Nonfar housholds lvng n th urban ara whr jatropha s producd (MNAJU); 3) Far housholds lvng n th rural ara whr jatropha s producd (MAJR); 4) Non-far 30 W hav basd ths dntfcaton on Pallèr t al. (2009).

19 housholds lvng n th rural ara whr jatropha s producd (MNAJR); 5) Far housholds lvng outsd th urban ara whr jatropha s producd (MANJU); 6) Non-far housholds lvng outsd th urban ara whr jatropha s producd (MNANJU); 7) Far housholds lvng outsd th rural ara whr jatropha s producd (MANJR); 8) Non-far housholds lvng outsd th rural ara whr jatropha s producd (MNANJR). Tabl 7 prsnts th shar of housholds as wll as avrag pr capta xpndtur for ach of th ght groups. W can obsrv that 10.15% of Malan housholds lv n th ara whr jatropha cultvaton s alrady actv. In addton, w can s that th avrag pr capta xpndtur of urban housholds, whthr lvng n th jatropha ara or not, s hghr than th natonal avrag, wth th xcpton of far housholds lvng n th urban jatropha ara (MAJU). Tabl 7: Dscrptv statstcs for groups GROUPS % of Man of pr capta housholds xpndtur (Fcfa) 1 MAJU Far housholds n urban jatropha ara 0,56% MNAJU Non-far housholds n urban jatropha ara 2,86% MAJR Far housholds n rural jatropha ara 4,93% MNAJR Non-far housholds n rural jatropha ara 1,48% MANJU Far housholds outsd urban jatropha ara 4,28% MNANJU Non-far housholds outsd urban jatropha ara 29,68% MANJR Far housholds outsd rural jatropha ara 40,51% MNANJR Non-far housholds outsd rural jatropha ara 15,71% MALI 100% Sourcs: El, 2006 Lt us bgn wth a brf analyss of povrty and nqualty for th rfrnc stuaton at th natonal lvl and for th ght groups of ntrst (Fgur 3). 31 Whl povrty at th natonal lvl s slghtly nfror to 40%, th jatropha ara s rvald to b th poorst n Mal for th whol of Malan housholds wth th xcpton of th group of non-far housholds lvng n urban aras (MNAJU), whch n fact provs to b th last poor of th groups (10.43%). Far housholds lvng n rural jatropha aras (MAJR) ar, convrsly, th ost affctd by povrty (51.27%). Fnally, t should b notd that th povrty of housholds lvng outsd th jatropha producton ara s systatcally wakr than natonal povrty. It wll b ntrstng to s whthr th sulatd polcs wll attnuat ths dffrnc. W should add that th sa rsults can b obsrvd for ndcs of dpth (FGT1) and svrty (FGT2). 31 Th wlfar ndcator usd s that of annual xpndturs followng prvat nt transfrs (ontary and non ontary), xprssd pr capta. Th ltratur shows that nco transfrs btwn housholds can b consdrd a sourc of povrty rducton as wll as a ans to attnuat nqualts btwn housholds (Adas and H, 1995). Coulbaly (2010) shows that n th absnc of transfrs, povrty would b uch hghr n Mal and ts rgons. It s also worth notng that th natonal thrshold chosn s th on suggstd by th World Bank, basd on th basc nds thodology. Th four othr thrsholds wr obtand usng avrags basd on rgonal urban and rural thrsholds whl takng nto account both jatropha producng aras and othrs. Tabl 10 of th appndx prsnts th thrsholds usd.

20 Mal provs to b a country wth sgnfcant nqualty (0.57). Th only group wth a hghr Gn ndx s that of non-far housholds lvng outsd th jatropha ara and n an urban sttng (MNANJR). Th two groups of non-far housholds lvng n an urban ara (MNAJU and MNANJU) and n whch th povrty ncdnc s th wakst (rspctvly 10.43% and 12.02%) ar n fact th groups arkd by strong nqualty (0.53). Th whol of groups contanng far housholds (MAJU, MAJR, MANJU and MANJR) hav Gn ndcs varyng btwn 0.43 and 0.47 blow natonal nqualty. Fnally, th group of non-far housholds lvng n th jatropha producton ara n a rural sttng (MNAJR) provs to b th last unqual. Fgur 3: Povrty and nqualty analyss for Mal and th groups (a) (b) Sourc: Rsutlts obtand by authors wth DASP 2.1 Wth rgard to th pact of sulatons on natonal povrty, w can s that Mal would xprnc a vry slght ncras n povrty whn th country would choos to ncras th quantty of land dvotd to jatropha producton (Tabl 8). Only a 20% subsdy n th boful sctor (S 4A) would hav a bnfcal ffct, albt a odst on, on povrty ncdnc (- 0.15%). Howvr, ths ffct s canclld whn th subsdy s accopand by an ncras n jatropha producton va an ncras n th land factor. Th provnt achvd wth th stablshnt of a subsdy n th boful sctor s a consqunc of th jatropha prx ncras of alost 15%, whl th othr sulatons nsnuat a dcras. Morovr, th gnralzd dcras n th prx of bofuls s uch wakr wth th subsdy polcy. Whn analyzng th pact of th polcs on th targtd groups, w obsrv that only two of ths groups xprnc changs n trs of povrty. 32 Th group of far housholds lvng n th urban jatropha ara provs to b th prary bnfcary of th ncras n land dstnd for jatropha producton. Indd, th ncdnc of povrty would dcras btwn 8.57% and 12.85%. Ths rsult s consstnt wth xpctatons snc th housholds wll b abl to xplot or land for jatropha producton, whch s confrd by th ncras n ts valu addd (Tabl 4). W ar, howvr, awar of th lts of ths rsult snc ths group of housholds s th sallst and rprsnts 0.56% of Malan housholds. 32 Th obtand rsults can b obsrvd for th thr povrty ndcs usd.

21 Tabl 8: Varaton of povrty ndcs by groups Bas S 1A S 1B S 2A S 2B S 3A S 3B S 4A S 4B Mal MAJU MNAJU MAJR MNAJR MANJU MNANJU MANJR MNANJR FGT0 39,72 46,09 10,43 51,27 35,73 20,09 12,02 38,59 34,93 FGT1 17,42 20,17 3,19 22,31 15,32 7,40 4,26 15,29 13,06 FGT2 9,96 11,30 1,22 12,06 8,11 4,03 2,24 8,17 6,61 Δ% FGT0 0,06-12,85 0,00 1,81 0,00 0,00 0,00 0,00-0,33 Δ% FGT1 0,28-12,26-0,05 5,45-0,02-0,02-0,03-0,02-0,03 Δ% FGT2 0,41-12,36-0,08 7,54-0,03-0,02-0,03-0,03-0,03 Δ% FGT0 0,06-12,85 0,00 1,81 0,00 0,00 0,00 0,00-0,33 Δ% FGT1 0,28-12,26-0,05 5,46-0,02-0,02-0,03-0,02-0,03 Δ% FGT2 0,41-12,36-0,08 7,55-0,03-0,02-0,03-0,03-0,03 Δ% FGT0 0,04-8,57 0,00 4,44 0,00 0,00 0,00 0,00 0,06 Δ% FGT1 0,22-12,04-0,56 6,76-0,28-0,16-0,33-0,24-0,19 Δ% FGT2 0,35-12,12-0,92 9,32-0,46-0,20-0,34-0,25-0,25 Δ% FGT0 0,03-12,85 0,00 1,81 0,00 0,00 0,00 0,00 0,06 Δ% FGT1 0,26-12,22-0,56 5,47-0,28-0,01-0,33 0,02-0,19 Δ% FGT2 0,37-12,30-0,92 7,56-0,46 0,00-0,34 0,03-0,25 Δ% FGT0 0,08-8,57 0,00 5,08 3,37 0,00 0,00 0,31-0,16 Δ% FGT1 0,14-11,76-1,07 7,95-0,55-0,30-0,63-0,46-0,37 Δ% FGT2 0,27-11,91-1,78 10,87-0,89-0,38-0,67-0,48-0,48 Δ% FGT0 0,04-12,85 0,00 1,13 3,37 0,00 0,00 0,44-0,16 Δ% FGT1 0,23-12,17-1,07 5,49-0,55 0,05-0,63-0,07-0,37 Δ% FGT2 0,33-12,23-1,78 7,56-0,89 0,02-0,67 0,08-0,48 Δ% FGT0-0,15-12,85 0,00-1,20 0,00 0,00 0,00 0,00-0,33 Δ% FGT1-0,21-13,29-0,12-1,65-0,07 0,73 0,02-0,03-0,01 Δ% FGT2-0,24-13,29-0,25-2,11-0,10 0,08-0,01-0,02-0,01 Δ% FGT0 0,03-12,85 0,00 1,20 0,00 0,00 0,00 0,00-0,33 Δ% FGT1 0,27-12,28-0,17 5,43-0,09 0,05-0,05-0,06-0,03 Δ% FGT2 0,39-12,34-0,32 7,51-0,13 0,05-0,02-0,05-0,04 Convrsly, th group of far housholds lvng n th rural part of th jatropha ara (MAJR) provs to b th losr whn t cos to sulatd asurs wth th xcpton, agan, of th subsdy polcy n th boful sctor. Th dcras n th prc of jatropha largly xplans ths rsult, snc far housholds s thr nco dnshd by ths prc dcras. Th last group for whch w obsrv slght provnts n povrty ndcs s that of non-far housholds lvng n rural aras but outsd th jatropha producton ara (MNANJR). Ths s xpland by th ncras obsrvd n pr capta xpnss obtand for ths houshold group. For th othr groups, th ffcts of th sulatons ar vrtually nxstnt, ncludng for sulaton 4A (subsdy n th boful sctor). It s nvrthlss ntrstng to obsrv that rgardlss of th sourc of ncras of land dstnd for jatropha producton, th ffct on povrty ncdnc s nl. Ths could ply that vn f thr s substtuton n agrcultural

22 producton btwn jatropha and anothr typ of agrcultur (as for sulatons 2 and 3), th housholds of ths groups would not b ngatvly affctd. 33 Tabl 9 prsnts th ffcts of th fv groups of sulatons on nqualts. Th sulatd polcs all hav a slghtly ngatv ffct on Malan nqualty. Thus, as nw lands dstnd for jatropha producton ar substtutd for lands usd for othr purposs (sulatons 1, 2 and 3), w obsrv an ncras n th Gn ndx. Evn f ths ncras s nscul (btwn 0.08% and 0.12% ncras), ths rsult ust b consdrd n th vntualty that asurs would b takn to proot jatropha producton n Mal. Tabl 9: Varaton of Gn ndx by groups Mal MAJU MNAJU MAJR MNAJR MANJU MNANJU MANJR MNANJR Gn 0,57 0,47 0,53 0,43 0,41 0,42 0,53 0,45 0,58 Δ% S 1A 0,077 11,088 0,002-0,255-0,001 0,000 0,003-0,004 0,002 Δ% S 2B 0,077 11,088 0,002-0,255-0,001 0,000 0,003-0,004 0,002 Δ% S 2A 0,091 10,918 0,027-0,292-0,010 0,041-0,010 0,004 0,015 Δ% S 2B 0,098 11,144 0,027-0,281-0,010 0,049-0,010 0,003 0,015 Δ% S 3A 0,103 10,783 0,053-0,318-0,018 0,083-0,018-0,008 0,030 Δ% S 3B 0,120 11,202 0,053-0,308-0,018 0,098-0,018-0,007 0,030 Δ% S 4A 0,024 12,414 0,000 0,158 0,001 0,005-0,004-0,016 0,007 Δ% S 4B 0,068 10,680 0,001-0,248 0,000 0,005-0,007-0,015 0,062 In trs of th ffcts on th groups, th rsults obtand ar consstnt wth thos found for povrty snc th two ost affctd groups ar far housholds lvng n th urban jatropha ara (MAJU) and far housholds lvng n a rural jatropha ara (MAJR). Th forr (MAJU) wll xprnc a growth of nqualty gratr than 10% rgardlss of th sulaton, thus approachng th nqualty obsrvd aong non-far housholds n ths ara (MNAJU). For far housholds lvng n th rural part of th jatropha ara (MAJR), th ncras n jatropha producton rsults n a slght dcras n nqualts. Only th prsnc of th subsdy (S 4A) would hav a ngatv ffct on th nco dstrbuton of ths housholds. As for povrty, th ffcts of th sulatons on nqualty ar vrtually nxstnt for th othr groups. 5.3 Pro-poor analyss To coplt our pact analyss, w hav xand whthr th suggstd sulatons could b consdrd by polcs gvn thr pro-poor natur. To conduct ths analyss, w hav usd th growth ncdnc curv (GIC) dvlopd by Ravallon and Chn (2003). Ths analyss allows us to s whthr th conoc growth gnratd by th odl s rdstrbutd to th poorst cntls. 33 Snstvty analyzs wll hav to b conductd n a rvsd vrson to support ths pont.

23 Fgur 4: Growth ncdnc curv for Mal (a) Sulaton 1B (b) Sulaton 4A Sourc: Coputd by authors fro ELIM 2006 wth DASP packag Fgur 4(a) prsnts ths GIC for Mal whn th quantty of land dstnd for jatropha producton was ultpld by 15 basd on non-cultvatd lands (S 1B). W should rcall that w had obsrvd an ncras n natonal povrty and an xacrbaton of povrty followng ths sulaton. GIC analyss shows us that ths asur would tnd to b prorch snc th frst cntls would xprnc a dcras n thr avrag nco whl th 50 last cntls would tnd to s an ncras n thr avrag nco. Whn w rplcat th analyss basd on th stablshnt of a subsdy n th boful sctor (S 4A), w obsrv that all cntls bnft fro an ncras n avrag nco. Housholds btwn th 40 th and 70 th cntls would b thos bnftng ost fro ths asur. In ths cas, t s nvrthlss dffcult to talk about a pro-poor or pro-rch asur. Fgur 5: Growth ncdnc curv for MAJR (a) Sulaton 1B (b) Sulaton 4A Sourc: Coputd by authors fro ELIM 2006 wth DASP packag Fgur 5 rprsnts th GIC for th group of far housholds lvng n th rural jatropha ara, whch s th poorst group of housholds (51.27%). Incrasng jatropha producton land ld to an ncras n povrty and a slght dcras n nqualts n ths group. Th GIC 34 shows us that ths asur has a ngatv ffct on all cntls but dos not allow us 34 W should add that vry slar curvs ar obtand for all sulatons wth th xcpton of sulaton 4A.

24 to conclud concrnng th pro-poor natur of th polcy. Only sulaton 4A shows an provnt n avrag nco for all housholds of th group. Nvrthlss, ths rsult dos not nabl us to thnk that ths asur would allow for rducng povrty aong th poorst of ths group. Ths analyss of pro-poor growth basd on asurs for ncrasng jatropha producton n Mal va ncrasng th land factor lads us to conclud that ths chans would not allow for rachng th objctv of rducng povrty. W hav found that subsdzng th boful sctor would b prfrabl at th natonal lvl as wll as for ost groups. Ths asur provs to b nsuffcnt, howvr, du to ts wak ffct on povrty and nqualts. Th analyss of pro-poor growth confrs that th ans would b ntrstng for th Malan conoy but ts xtnt would b nsuffcnt. 6 Concluson and rcoandatons In th papr, w construct a CGE odl wth a crosulaton odl to analyz th dstrbutonal pact of th xpanson of th jatropha and boful sctors n Mal. W also provd acro and sctor ffcts of such scnaros. Gvn th sall ntal sz of th sctor, w show that a sgnfcant xpanson of th sctor wll not produc a ajor ngatv pact on othr agrcultural sctors. Our prlnary rsults also s to show that th xpanson of th sctor s not vry ffcnt to achv th goal of rducng povrty. So groups drctly concrnd s to bnft but othr ar losrs and povrty ncrass at th natonal lvl. Th only scnaro producng postv rsults at th natonal lvl s th subsdy to boful fundd by a rducton n th producton of publc srvcs. It s also ntrstng to obsrv that th orgn of th land usd for xpndng producton s not a ky lnt to contrbut to our fndngs. Hnc, usng dl land or coptng wth othr sctors producs a slar povrty pact. Furthr scnaros wll nd to b analyzd and snstvty analyss prford to s how robust ths rsults ar to changs n closur of th currnt account balanc and so ky lastcts n our odl, such as th dgr of substtuton btwn nputs n producton, trad lastcts. In concludng w wll phasz th portanc of such an analyss n a CGE crosulaton contxt vn wth th rlatvly sall sz of th jatropha and boful sctors. Our rsults rval that polcy scnaros prford on two sctors that contrbut lss than 0.5% can produc a sgnfcant dstrbutonal pact. Th lnks btwn ths sctors and th rst of th conoy allow us to dntfy ntrrlatons btwn a scnaro and povrty and nco dstrbuton changs. Furthr analyss s nvrthlss rqurd to ncras th confdnc w can hav n ths conclusons.

25 7 Bblography Achtn, W. M. J., L. Vrchot, Y. J. Frankn, E. Mathjs, V. P. Sngh, R. Arts and B. Muys, (2008), Jatropha bo-dsl producton and us, Boass and Bonrgy, Vol 32(12), pp Adas, R. H. Jr. and J. J. H, (1995), "Sourcs of nco nqualty and povrty n rural Pakstan", Rsarch rports 102, Intrnatonal Food Polcy Rsarch Insttut (IFPRI). Bar, S. L., M. Clnts, C. W. Grffths, and J. E. Ihrg, (2009), Bofuls Ipact on Crop and Food Prcs: Usng an Intractv Spradsht, FRB Intrnatonal Fnanc Dscusson Papr No Avalabl at SSRN: Bng, M., Assssnt of th potntal of Jatropha curcas, (bodsl tr) for nrgy producton and othr uss n dvlopng countrs, ECHO s wbst: Avalabl at td68. Boccanfuso, D. and C. Ménard, (2008), La crossanc pro-pauvr : un aprçu xhaustf d la «bot à outls», Cahr d rchrch du GRÉDI #09-06, Unvrsté d Shrbrook, Québc, Canada. Bourgugnon F., A. S. Robllard, and S. Robnson. (2005). Rprsntatv Vrsus Ral Housholds n th Macroconoc Modlng of Inqualty. In T. J. Kho, T.N. Srnvasan, and J. Whally (ds.), Frontrs n Appld Gnral Equlbru Modlng. Cabrdg: Cabrdg Unvrsty Prss. Bourgugnon, F. and A. Spadaro. (2005). Mcrosulaton as a Tool for Evaluatng Rdstrbuton Polcs. Journal of Econoc Inqualty, 4(1): Bourgugnon F. and L. Savard (2008). Dstrbutonal Effcts of Trad Rfor: An Intgratd Macro- Mcro Modl Appld to th Phlppns n Bourgugnon, F., M. Bussolo, and L.A. Prra da Slva, (ds.) Th Ipact of Macroconoc Polcs on Povrty and Inco Dstrbuton: Macro-Mcro Brttan R. and N. Lutalado, (2010), Jatropha: A Sallholdr Bonrgy Crop - Th Potntal for Pro-Poor Dvlopnt, FAO Intgratd Crop Managnt Vol. 8, 116 pg ISBN: Avalabl at: Capa, C. and C. Valntn, (2009), L jatropha, un plant racl? n Scncs au Sud n 49 Trbun. Avalabl at: Chn, S. and M. Ravallon. (2004). Wlfar Ipacts of Chna s Accsson to th World Trad Organzaton. World Bank Econoc Rvw 18(1): Cockburn, J. (2001). Trad Lbralzaton and Povrty n Npal: A Coputabl Gnral Equlbru Mcrosulaton Analyss. Workng papr 01-18, CREFA, Unvrsty of Laval, Québc. Cognau, D. and A.-S. Robllard. (2000). Inco Dstrbuton, Povrty and Growth n Madagascar: Mcro Sulatons n a Gnral Equlbru Frawork. IFPRI TMD Dscusson Papr 61, Intrnatonal Food Polcy Rsarch Insttut, Washngton, DC. Coulbaly M., (2010), Crossanc éconoqu t dynaqu d la pauvrté, Cahr du GREAT, Baako. Dcaluwé, B. J.-C. Duont, and L. Savard. (1999). How to Masur Povrty and Inqualty n Gnral Equlbru Frawork. Workng Papr 9920, CREFA, Unvrsty of Laval, Québc. Dcaluwé, B., A. Martns, and L. Savard. (2001a). La poltqu conoqu du dévloppnt t ls odèls d qulbr général calculabl. Montréal: Montréal Unvrsty Prss. Dornbosch R. and R. Stnblk (2007), Bofuls: Is th cur wors than th dsas? OECD, Round Tabl on Sustanabl Dvlopnt, Pars, Sptbr 2007, [ SG/SD/RT(2007)3/REV1], Organsaton for Econoc Dvlopnt and Coopraton (OECD), Pars. Fargon, F.; Hll, J.; Tlan, D.; Polasky, S. and Hawthorn, P., (2008), Land clarng and th boful carbon dbt, Scnc Vol. 319, pp

26 Fostr (J.), (J.) Grr t (E.) Thorbck, «A class of Dcoposabl Povrty Masurs», Econotrca, 52(3). GEXSI, (2008), Global Markt Study on Jatropha. Fnal Rport. Prpard for th World Wd Fund for Natur (WWF). London/Brln: Global Exchang for Socal Invstnt.Avalabl at: Gørtz, M., G. Harrson, C. Nlsn, and T. Ruthrford. (2000). Wlfar gans of xtndng opnng hours n Dnark. Econoc workng papr B-00-03, Unvrsty of South Carolna, Darla Moor School of Busnss, Coluba, South Carolna. Haïdara, A., O., (1996), Valorsaton d un hul végétal tropcal : l hul d pourghèr», Méor d aîtrs s scncs applqués, Gén Chqu, Unvrsté d Shrbrook. Avalabl at: Hady D. and S. Fan, (2008), Anatoy of a crss: th causs and consquncs of surgng food prcs, Agrcultural Econocs, Vol. 39, supplnt, pp Hnnng, R.,K., (2009), Th Jatropha Syst an ntgratd approach of rural dvlopnt, Wssnsbrg, Grany. Avalabl at: Hrtl, T., and J. Rr. (2005). "Prdctng th povrty pacts of trad rfor, Journal of Intrnatonal Trad & Econoc Dvlopnt, 14(4), pp Hll, J., E. Nlson, D. Tlan, S. Polasky, and D. Tffany. (2006), Envronntal, conoc, and nrgtc costs and bnfts of bodsl and thanol bofuls, Procdngs of th Natonal Acady of Scnc 103: Hupp, M., and M. Ravallon. (1991). Th Sctoral Structur of Povrty Durng an Adjustnt Prod: Evdnc for Indonsa n th Md-1980s. World Dvlopnt 19(2): Kng, D., and Handa S. (2003). Th wlfar ffcts of balanc of paynts rfors: A acro-cro sulaton of th cost of rnt-skng.? Th Journal of Dvlopnt Studs, 39(3): Mathws, J., (2007), Prospcts for a bofuls ndustry n Australa, MGSM workng paprs n anagnt, Macquar Unvrsty. Graduat School of Managnt, no. 13. Mtchll, D., (2008), A Not on rsng food prcs, Th World Bank, Dvlopnt Prospcts Group, Polcy rsarch workng papr #4682. Latap, R., (2007), "La cultur du pourghèr : un actvté génératrc d rvnus qu prt d far fac aux njux énrgétqus du Mal. L cas du projt Garalo Bagan Yln", Unvrsté d Rnns 1 t MalFolkcntr, 107 pags. Avalabl at: Lgndr, B., (2008), Jatropha Curcas (Tabanan) : Not agronoqu, Tchnologs for han dvlopnt, Dakar, Sngal. Avalabl at: Luoa J. R., (2009), Hald as a Mracl Boful, Jatropha Falls Short of Hyp, Yal Envronnt 360, publcaton of th Yal School of Forstry & Envronntal Studs. Avalabl at: Rsng Food Prcs, Polcy Rsarch Workng Papr No. 4682, Th World Bank, Washngton, DC. OECD, (2008), Boful Support Polcs: An Econoc Assssnt, Avalabl at n12/contp1-1.ht. OECD-FAO, (2009), Agrcultural Outlook , 276 pags, Avalabl at:

27 Pallèr, G. and S. Fauvaud, (2009), «Ls njux ds agrocarburants pour l ond paysan au Mal», GERES, AMEDD, ICD, Avalabl at : Razon, D., (2009), «Cultur ntnsv d Jatropha : Mann ou catastroph?», RFI, In ln on 23/06/2009, Ravallon, M., (1994), «Povrty Coparsons», Harwood acadc publshrs, 143 pags. Ravallon, M. and S. Chn (2003). "Masurng Pro-Poor Growth", World Bank, Polcy Rsarch Workng Papr Républqu du Mal, (2008), Orntatons stratégqus t prortés d'nvstssnts agrcols», Conférnc d Tabl Rond ds ballurs d fonds du Mal, Baako, Rnnr, R., (2007), Grn Gold n a Shrub: Entrprnurs targt th jatropha plant as th nxt bg boful, Grn Gold n a Shrub, Scntfc Arcan, Jun. Avalabl at: Ruthrford, T., D. Tarr, and O. Shpotylo. (2005). Povrty Effcts of Russa s WTO Accsson: Modlng Ral Houshold and Endognous Productvty Effcts. World Bank Polcy Rsarch Workng Papr 3473, Washngton, DC. Savard, L. (2005). Povrty and Inqualty Analyss wthn a CGE Frawork: A Coparatv Analyss of th Rprsntatv Agnt and Mcrosulaton Approachs. Dvlopnt Polcy Rvw 23(3): Sarchngr, T.; Hlch, R.; Houghton, R.A.; Dong, F.; Elobd, A.; Fabosa, J.; Tokgoz, S.; Hays, D. and Yu, T. (2008), Us of U.S. croplands for bofuls ncrass grnhous gass through ssons fro land us chang, Scnc Vol. 319, pp Vos, R, D Jong N. (2003) Trad lbralzaton and povrty n Ecuador: a CGE acrocrosulaton analyss. Econoc Syst Rsarch, 15(2): Word Bank, (2007), Mal - Scond povrty rducton stratgy papr and jont IDA-IMF staff advsory not, Rport # 40724, 241 pags. Avalabl at /Rndrd/PDF/407240ML0IDA1ScM an.pdf.

28 8 Appndx Fgur 6: Th uss of Jatropha curcas and th nrgy valus of ts coponnts Sourc: Brttan R. and N. Lutalado (2010) a Enrgy valus of th coponnts ar gvn n MJ/ kg Fgur 7: Prncpal jatropha cultvaton sts n Mal Sourc: Pallèr t al. (2009)