Macro-Micro Feedback Links of Irrigation Water Management in Turkey

Transcription

1 Publi Dislosure Authorized Publi Dislosure Authorized Publi Dislosure Authorized Publi Dislosure Authorized Pol i y Re s e r h Wo r k i n g P p e r 4781 Mro-Miro Feedbk Links of Irrigtion Wter Mngement in Turkey Erol H. Ckmk Hsn Dudu Sirin Sroglu Xinshen Dio Terry Roe Yov Tsur The World Bnk Development Reserh Group Sustinble Rurl nd Urbn Development Tem November 2008 WPS4781

2 Poliy Reserh Working Pper 4781 Abstrt Agriulturl prodution is hevily dependent on wter vilbility in Turkey, where hlf the rop prodution relies on irrigtion. Irrigted griulture onsumes bout 75 perent of totl wter used, whih is bout 30 perent of renewble wter vilbility. This study nlyzes the likely effets of inresed ompetition for wter resoures nd hnges in the Turkish eonomy. The nlysis uses n eonomy-wide Wlrsin Computble Generl Equilibrium model with detiled ount of the griulturl setor. The study investigted the eonomywide effets of two externl shoks, nmely permnent inrese in the world pries of griulturl ommodities nd limte hnge, long with the impt of the domesti rellotion of wter between griulturl nd non-griulturl uses. It ws lso reognized tht beuse of sptil heterogeneity of the limte, the simulted senrios hve differentil impt on the griulturl prodution nd hene on the llotion of ftors of prodution inluding wter. The gretest effets on mjor mroeonomi inditors our in the limte hnge simultions. As result of the trnsfer of wter from rurl to urbn res, overll prodution of ll rops delines. Although prodution on rinfed lnd inreses, prodution on irrigted lnd delines, most notbly the prodution of mize nd fruits. The derese in griulturl prodution, oupled with the domesti prie inrese, is further refleted in net trde. Agriulturl imports inrese with greter deline in griulturl exports. This pper produt of the Sustinble Rurl nd Urbn Development Tem, Development Reserh Group is prt of lrger effort in the deprtment to minstrem reserh on role of wter resoures in the eonomy. Poliy Reserh Working Ppers re lso posted on the Web t The uthors my be ontted t kmke@ metu.edu.tr, dudu@metu.edu.tr, ssirin@metu.edu.tr, x.dio@gir.org, troe@umn.edu, tsur@gri.huji..il. The Poliy Reserh Working Pper Series dissemintes the findings of work in progress to enourge the exhnge of ides bout development issues. An objetive of the series is to get the findings out quikly, even if the presenttions re less thn fully polished. The ppers rry the nmes of the uthors nd should be ited ordingly. The findings, interprettions, nd onlusions expressed in this pper re entirely those of the uthors. They do not neessrily represent the views of the Interntionl Bnk for Reonstrution nd Development/World Bnk nd its ffilited orgniztions, or those of the Exeutive Diretors of the World Bnk or the governments they represent. Produed by the Reserh Support Tem

3 MACRO-MICRO FEEDBACK LINKS OF IRRIGATION WATER MANAGEMENT IN TURKEY * Erol H. Ckmk, Hsn Dudu **, Sirin Sroglu **, Xinshen Dio ***, Terry Roe ****, Yov Tsur ***** JEL lssifition: C68; O13 ; Q15; Q18 Key words: Computble Generl Equilibrium; Feedbk links; Irrigtion Wter; Turkey * This pper is produt of the projet Mro-Miro Feedbk Links of Irrigtion Wter Mngement, funded by the Reserh Committee of World Bnk nd mnged by DECRG-RU. We re indebted to Ariel Dinr dvie nd omments. Construtive review omments from Erin Yeldn re knowledged with ppreition. Ckmk, Dudu, Sroglu, Roe nd Tsur were onsultns to the World Bnk on this study. ** Deprtment of Eonomis, Middle Est Tehnil University (METU), Ankr, Turkey. *** Interntionl Food Poliy Reserh Institute (IFPRI), Wshington, DC, USA **** Deprtment of Applied Eonomis, University of Minnesot, St. Pul, MN, USA. ***** Deprtment of Agriulturl Eonomis nd Mngement, Hebrew University of Jeruslem, Rehovot, Isrel.

4 Aronyms ARIP: Agriulturl Reform Implementtion Projet CES: Constnt Elstiity of Substitution CSE: Consumer Subsidy Equivlents CGE: Computble Generl Equilibrium DIS: Diret Inome Support DSI: Stte Hydruli Works, Republi of Turkey EU: Europen Union FAO: Food nd Agriulture Orgniztion IFPRI: Interntionl Food Poliy Reserh Institute GDRS: Generl Diretorte of Rurl Servies GDP: Gross Domesti Produt GTAP: Globl Trde Anlysis Projet LSCB: Lower Seyhn-Ceyhn Bsin OECD: Orgniztion for Eonomi Co-opertion nd Development PSE: Produer Subsidy Equivlents QHS: Quntittive Household Survey REC: Regionl Environment Center SAM: Soil Aounting Mtrix SPO: Stte Plnning Orgniztion, Republi of Turkey TACOGEM-W: Turkish Agriulturl Computble Generl Equilibrium Model with wter TurkSTAT: Turkish Sttistil Institute UFT: Underseretrit of Foreign Trde, Republi of Turkey VA: Vlue-dded WUA: Wter User Assoitions WTO: World Trde Orgniztion 2

5 Tble of Contents Key words: Computble Generl Equilibrium; Feedbk links; Irrigtion Wter; Turkey... 1 Aronyms... 2 Tble of Contents... 3 List of Tbles... 4 List of Figures... 5 List of Figures Introdution Wter nd Agriulturl Setors in Turkey Wter Resoure Avilbility nd Use in Turkey Overview of the Agriulturl Setor in Turkey The Modeling Frmework Struture of the Agriulturl Computble Equilibrium Model for Turkey The Frm Model: Struture nd Results Linkge between the Frm Model nd TACOGEM-W Empiril Results Aggregte Results of the Simultions Senrio 1: Inrese in World Pries Senrio 2: The Urbniztion Senrio Senrio 3: Climte Chnge Senrio Conlusions, Poliy Implitions, nd Future Reserh Agend...57 Referenes Appendix A: Additionl Tbles Appendix B: The Algebri Struture of the CGE Model Appendix C: The Algebri Struture of the Frm Model

6 List of Tbles Tble 2.1. Turkey: Wter Resoures Potentil nd Use...9 Tble 2.2. Wter Potentil nd Lnd Distribution by Bsins...11 Tble 2.3. Setorl Wter Use in Turkey...12 Tble 2.4. Irrigtion Development by Regions, 2007 (1,000 h)...13 Tble 2.5. Irrigtion rtios of the res trnsferred by DSI, Tble 2.6. Mngement of Irrigtion Shemes, Tble 2.7. Agriulturl vlue-dded: growth nd shre, Tble 3.1. Gross ntionl produt by kind of tivity, Tble 3.2. Employed persons by brnh of eonomi tivity, Tble 3.3.Shre of regionl prodution vlue in ntionl prodution vlue, 2003 (Agriulture only)...25 Tble 3.4. Agriulturl lnd use, Tble 3.5. Irrigted versus rinfed lnd use, 2001 (hetres)...26 Tble 3.6. Irrigted re by irrigtion soure, Tble 3.7. Irrigted re by irrigtion system, Tble 4.1. Effets on GDP...35 Tble 4.2. Mroeonomi results of simultions...36 Tble 4.3. Prie Shoks in Senrio Tble 4.4. Chnge in domesti verge output pries nd omposite good supply nd domesti prodution...38 Tble 4.5. Perentge hnge in shre of exports nd imports in totl prodution...39 Tble 4.6. Perentge hnge in ftor employment...40 Tble 4.7: Chnge in ftor pries...41 Tble 4.8.Urbn popultion dynmis in Turkey...42 Tble 4.9. NUTS Level-1 Regions, net internl migrtion ( )...44 Tble Chnge in totl prodution quntity...46 Tble Chnge in ftor pries...47 Tble Chnge in export quntities...49 Tble Import quntity nd vlue...50 Tble Chnge in prodution of irrigted tivities...54 Tble 4.15.Chnge in ftor pries...55 Tble Chnge in disggregted rel household onsumption...56 Tble 5.1. Summry Impt Mtrix...58 Tble A.1. NUTS Regions...67 Tble A.2. Model Regions...68 Tble A.3. Aggregted SAM for Turkey, 2003 (Billion TL)

7 List of Figures Figure 3.1. The five regions defined in the SAM Figure 3.2. Shdow rent Figure 4.1: In-migrtion by ples of residene (%), Figure 4.2: Annul totl preipittion (mm), Adn Figure 4.3: Annul totl preipittion (mm), Diyrbkır Figure 4.4: Annul totl preipittion (mm), Kony

8 1 Introdution The griulturl setor is n importnt setor in the Turkish eonomy. This setor is mjor soure of employment ounting for 27 perent of the totl workfore, nd providing employment for pproximtely 70 perent of the rurl workfore. However, similr to other rpidly growing eonomies, the shre of griulture in Turkey s GDP hs delined from 30 perent in the lte 1970s to 9 perent in The griulturl setor, overll, ppers to lg the rest of the eonomy in trnsforming to one with omprble per pit inomes. The growth rte of griulturl vlue-dded is bout one-fourth of the rest of the eonomy, whih explins the delining shre of griulture in GDP over the pst three dedes. Irrigtion hs signifint role in griulturl prodution. Irrigted griulture forms bout hlf of the rop prodution vlue. Diverse limti zones, rnging from Mediterrnen to semi-rid ontinentl limte, nd vried regionl vilbility of wter resoures in Turkey, imply tht wter is mjor ftor in inresing produtivity nd deresing voltility in griulturl prodution. Almost ll export oriented rops (fruits nd vegetbles) nd importompeting rops (otton nd mize) re hevily dependent on irrigtion. Development of wter infrstruture gined momentum in the lte 1960s. Irrigted re hs more thn doubled sine then nd the storge pity of dms rehed 140 billion m 3. Due to n inrese in popultion, per pit wter vilbility is down to roughly 1,700 m 3 per yer. The irrigtion setor is urrently onsuming bout 75 perent of totl wter onsumption whih orresponds to bout 30 perent of renewble wter vilbility. The non-griulturl demnd for wter is inresing rpidly due to the fst pe of urbniztion nd industriliztion. However, supply-mngement prties trgeting minly the development of irrigtion infrstruture, hve ontinued to previl s the mjor determinnt of irrigted griulture. Even with the rpid expnsion of the irrigted re, rehing bout 20 perent of the totl re suitble for irrigtion, the growth of the griulturl vlue-dded hs been disml. The verge growth rte is only 1.3 perent per nnum in the pst four dedes, whih is lower thn the nnul popultion growth rte. Turkey hieved lmost full liberliztion of trde in the mnufturing setor in the 1990s. High protetion in griulture hs been mintined in order to sustin the self-suffiieny in mjor stples. As result, trnsfers to griulture tht re provided minly through prie distortionry mesures, rehed 3-4 perent of GDP. This poliy setting hs 6

9 prevented mjor struturl hnges in griulture. The ineffetive set of poliy tools nd their inresing burden on government expenditures led griulturl subsidiztion poliies to undergo mjor hnges in the ltest struturl djustment nd stbiliztion progrm t the strt of the new millennium. The mroeonomi stbiliztion progrm, inorporting tight fisl nd monetry poliies, hs ffeted the wter relted fetures of the Turkish griulturl setor. The griulturl subsidiztion reform progrm ws never trnsformed into reform in griulturl poliies involving the wter setor. A lrge number of lredy plnned, yet inomplete irrigtion development projets hve been further delyed due to lk of investble funds. The previling irrigtion poliy frmework for infrstruture development hs been revitlized s the eonomy reently reuperted, with the help of the progrm, without ny emphsis on the more effiient use of wter resoures. However, severl developments will ertinly inrese the ompetition for wter resoures nd my stipulte rdil hnges in wter poliies in the medium nd long run. The rpid pe of urbniztion my led to hnges in the inter-setorl llotion of wter. The Mediterrnen bsin is expeted to be severely ffeted by the limte hnge. This my further inrese the pressure on n lredy stressed wter eonomy with severe implitions for the griulturl setor. Growing interest by developed ountries in biofuels, ombined with inresing energy pries, is ggrvting the impt of limte hnge. Although reent surges in pries of bsi stples hve begun to deline, the griulturl ommodity pries, however, re expeted to remin high ompred to historil verges. In ddition, the renewl of the WTO- Agreement on Agriulture nd Turkey s ndidy for membership to the Europen Union (EU) will dd new dimensions to the delibertions on griulturl nd wter poliies. Bsed on this bkground it is therefore neessry to evlute the onsequenes of reent hnges in the ntionl nd interntionl senes in the Turkish eonomy to provide better dilogue with poliy mkers nd to develop proper poliy responses. The purpose of this study is to nlyze the potentil effets of surging griulturl pries, limte hnge, nd urbniztion in the Turkish eonomy by using n eonomy-wide model. The Wlrsin CGE model for Turkey disggregtes the eonomy into 20 griulturl nd 9 nongriulturl tivities. The griulturl setor is further disggregted into 5 regions. The model inorportes griulturl nd non-griulturl wter use with the differentited irrigted nd 7

10 rinfed griulturl prodution tivities t the regionl level. Furthermore, frm level model is used to estimte the shdow vlue of wter in griulture. The following setion presents n overview of wter nd griulturl setors, followed by brief review of relevnt CGE pplitions. Setion three provides desription of the CGE model used long with summry of mjor dt soures. This setion lso inludes the frm level model used in the estimtion of shdow pries for irrigtion wter. Vrious senrios onduted with the CGE model re explined nd the results re disussed in the fourth setion. The finl setion is reserved for the onlusions nd implitions of the of the study. 2 Wter nd Agriulturl Setors in Turkey Import substituting griulturl ommodities (mize nd otton) nd most of the exported produts (fruits nd vegetbles) re irrigtion intensive. Although Turkey is urrently using only 40 perent of its vilble wter resoures, it is estimted tht the ountry will reh its limits of vilble wter within two dedes due to inresing demnds from ll setors. The estimtes re bsed on the inrese in the non-griulturl demnd nd the full development of 8.5 million hetres of eonomilly irrigble lnd. Supply-side wter poliies still dominte, despite the widely pronouned pressures from the demnd side. Over-bstrtion of groundwter in some regions nd over-use of surfe wter in others ontinue to be the mjor supply side issues in the wter setor. Due to the frequent eonomi rises from the pst two dedes, nd the mismngement of griulturl poliies, struturl hnge in Turkish griulture hs been delyed. The setor still domintes the rurl eonomy, providing bout 70 perent of totl rurl employment. The dulisti struture of prodution hs ll the bsi trits of developing eonomy. It hs dominnt shre of prodution onentrted in smll holdings, while o-existing with ommeril nd mostly export-oriented produers. 2.1 Wter Resoure Avilbility nd Use in Turkey Turkey s limte is moderted by both the Mediterrnen nd ontinentl wether ptterns whih displys geo-limti diversity when ombined with highly vried topogrphy. The verge nnul temperture is o C on the southern ost, o C on the west ost, nd flututes between 4 to 19 o C in the interior regions, depending on their distne from se level. The nnul 8

11 verge preipittion is 643 mm, yet vries from 250 mm in the entrl prt to 3000 mm in the Estern Blk Se region. Seventy-five perent of nnul rin flls during the winter seson. Annul rinfll is less thn 500 mm in the inlnd Thre nd in the Estern Antoli regions. This diverse preipittion struture emphsizes the ruil importne of irrigtion. Generlly, griulturl prodution is dversely ffeted by the shortge nd inonsisteny of rinfll during the growing seson. Solr energy mkes it possible to grow rid nd semi-rid rops suh s bnns nd itrus. Moreover, it is possible to grow 2 to 3 different rops in irrigted res tht hve rop growing sesons for period of 270 dys. However, some rops my be hrvested before mturtion, prtiulrly in Estern Antoli with its 60 to 90 growing dys. The southest region hs very low humidity level. The ostl regions re humid with high preipittion rtes. Inevitbly, the topogrphi fetures re min ftors shping the distribution. The long-term nnul evportion rtes indite high rte, prtiulrly in the southest region, whih reeives lmost no rinfll during the summer, nd rehes more thn 2000 mm per yer in the Southestern region (Knber et l., 2005). The verge nnul preipittion of the ountry orresponds to wter potentil of 501 km 3 per yer, of whih 274 km 3 re lost to evpotrnspirtion, 69 km 3 feed quifers nd 158 km 3 flow through the rivers to the se or lkes. The gross totl surfe nd ground wter potentil of Turkey mounts to 234 km 3 (Tble 2.1). Tble 2.1. Turkey: Wter Resoures Potentil nd Use Surfe wter Groundwter Totl Surfe flow 158 km 3 Feeding groundwter 69 km 3 Men nnul preipittion 501 km 3 (603mm) Surfe runoff 193 km 3 Rehrge 41 km 3 Renewble wter potentil 234 km 3 Usble surfe runoff b 98 km 3 Sfe yield 14 km 3 Usble (net) 112 km 3 Consumption Consumption Consumption Notes: inludes the ontributions from underground (28 km 3 ) nd from the neighboring ountries (7 km 3 ); b inludes the usble flow of 3 km 3 from the neighboring ountries. Soures: DSI, The mount of surfe wter utilized for onsumption purposes is in the rnge of 98 km 3 per yer, inluding the ontributions from the neighboring ountries. Aording to the studies bsed on groundwter resoures, the totl sfe yield of groundwter resoures is estimted to be 9

12 14 km 3. Thus, the totl potentil vilble wter resoures from surfe flow nd groundwter would mount to 112 km 3 per yer. The ountry s surfe runoff is unevenly distributed in both time nd ple, onsistent with preipittion. Surfe nd ground wter resoures re limited in the Aegen, Thre nd Centrl Antoli regions where the demnd for wter is higher thn the rest of Turkey. The Aegen nd Thre Regions re highly urbnized nd industrilized, nd hve soil resoures suitble for irrigtion. They hve 10.5 perent of totl surfe wter resoures for the ountry while overing 19.3 perent of the entire re. Almost 30 perent of the totl surfe wter for the ountry flows through two rivers, the Tigris nd Euphrtes (Tble 2.2). An irregulr regime of rivers requires reservoirs to regulte the wter. It is estimted tht 98 km 3 of surfe wter (51 perent of totl surfe wter) n be onsumed by tehnilly nd eonomilly fesible projets. The tul utilizble wter mount in Turkey is round 1,700 um/person/yer in

13 Tble 2.2. Wter Potentil nd Lnd Distribution by Bsins Annul Bsin Effiieny (l/s/km²) Averge Annul Flow (km 3 ) Frming Lnd (1,000 h) Irrigble Lnd (1,000 h) Are No NAME (km²) 1 Mritz-Ergene 14, , , Mrmr 24, Susurluk 22, North Aegen 10, Gediz 18, K.Menderes 6, B.Menderes 24, , West 20, Mediterrnen 9 Antly 19, Burdur Lkes 6, Akry 7, Skry 58, , , West Blk Se 29, Yeşilırmk 36, , , Kızılırmk 78, , , Kony inlnd 53, , , Est 22, Mediterrnen 18 Seyhn 20, Orontes 7, Ceyhn 21, Euphrtes 127, , , Est Blk Se 24, Çoruh 19, Ars 27, Lke Vn 19, Tigris 57, , ,137.6 TOTAL 779, , ,753.6 Soure: DSI, Setorl onsumption of wter is presented in Tble 2.3. Totl humn nd utility wter onsumption is inresing stedily with popultion nd inome growth, totling 6.2 km 3 per nnum in The shre of the popultion served by dequte wter from the network onneted t home or stndpipes, rehed 85 perent in rurl res, nd 98 perent in urbn res. Annul wter lloted to industry is bout 4.1 billion m 3 supplied minly from groundwter resoures. 11

14 Tble 2.3. Setorl Wter Use in Turkey Irrigtion Domesti Industry Use (hm 3 ) 22,016 29,300 29,600 72,000 % of totl Use (hm 3 ) 5,141 5,800 6,200 18,000 % of totl Use (hm 3 ) 3,443 4,200 4,300 22,000 % of totl Totl use (hm 3 ) 30,600 39,300 40, ,000 Note: Both the setorl nd totl use vries depending on the soure. We hve presented the most reent findings. Trget mentioned in DSI (2007) implies full utiliztion of ll usble wter supplies. Soures: SPO, 2007; DSI, The totl irrigted re ws 5 million hetres in 2007 (Tble 2.4) with 75 perent of the wter lloted to irrigtion. The irrigted re hs lredy rehed 60 perent of the totl eonomilly irrigble re of 8.5 million hetres. Wter onsumption per hetre mounts to more thn 6,000 m 3. With respet to geogrphil regions used in this study, more thn 40 perent of the totl irrigted re is loted in the Western region. 12

15 Tble 2.4. Irrigtion Development by Regions, 2007 (1,000 h) DSI Region Geo.R DSI DSI (GWIC) GDRS Frmers Totl 1 Burs W Izmir W Eskisehir W Kony C Ankr C Adn W Smsun C Erzurum E Elzig E Diyrbkir SE Edirne W Kyseri C Antly W Istnbul W Snliurf SE Vn E Isprt W Sivs C K.Mrs SE Aydin W Trbzon E Kstmonu C Krs E Blikesir W Artvin E Totl 2, ,394 1,034 5,001 Soures: DSI, 2008b ; GDRS, 2007 ; SPO, The Western region is more populted nd industrilized ompred to the rest of the ountry. In ddition, the seven river bsins in this re re estimted to hve lredy exeeded their long-term pity utiliztion rtes (World Bnk, 2007). About 90 perent of irrigtion methods depend on grvity systems with low wter effiieny. Furthermore, the irrigtion rtios of the shemes trnsferred by Stte Hydruli Works (DSI) indite tht bout 35 perent of the irrigted re re lloted to rinfed griulture (Tble 2.5). Unvilbility of wter is the top reson for the shift to rinfed griulture (DSI, 2008). 13

16 Tble 2.5. Irrigtion rtios of the res trnsferred by DSI, Trnsferred Are Irrigtion Rtio Totl Irrigtion Rtio (1,000 h) (%) (%) , , , , , , , , Note: Inludes the re irrigted by frmers nd the re outside the Sheme Soure: DSI, 2008b. The signifint role of irrigtion for improving the performne of the griulturl setor is reognized in the Ninth Ntionl Development Pln for (SPO, 2007). The Pln estblishes the priority for the effiient use of wter resoures. The development of irrigtion infrstruture tkes preedene ording to the trgeted development towrd The trnsfer of opertion nd mintenne of the shemes developed by DSI is omplete (Tble 2.6). However, the sustinbility of trnsferred shemes to the benefiiry is questioned in the reent development pln. The Pln reommends to develop prtiiptory mehnisms together with the neessry legl provisions for effiient nd sustinble use of soil nd wter resoures. Tble 2.6. Mngement of Irrigtion Shemes, 2007 Number Are (1,000 h) Developed by DSI ,573 Mnged by DSI Trnsferred to users orgniztions 697 2,053 Trnsferred to ICs 1, Developed by GDRS nd mnged by the n. 1,394 frmers Developed nd mnged by the frmers n. 1,034 Totl n. 5,001 Soure: DSI (2008) Speifi legisltion regrding the wter setor with omprehensive wter mngement nd defining wter rights does not exist. Efforts in the pst ouple of yers to ent wter lw hve yet to be suessful. Setorl priorities re set in the DSI lw, but ler definitions of wter 14

17 entitlements re bsent. DSI s priority list begins with drinking nd industril wter supply, ontinues with irrigtion, power genertion, flood ontrol, nd ends with nvigtion. Priing nd ost reovery poliies vry mong setors. There is lmost no volumetri system for irrigtion, wheres volumetri hrges re ommon in domesti nd industril use. Almost ll wter users orgniztions determine the per hetre fee for the opertion nd mintenne bsed on expeted opertion nd mintenne osts. The government hs been relutnt to reuperte the investment osts. Severl legisltions nd regultions ddress speifi issues, but they re fr from forming n integrted frmework for effetive mngement of wter resoures. The existing lws nd regultions re lso fr from defining pproprite wter rights. Extended drought periods resulted in the full development of wter resoures in the western nd entrl regions involving the trnsfer of wter from irrigtion to domesti nd industril use. This sitution will inrese the unertinty of irrigtion wter llotion dversely ffeting frmers welfre. The legisltive rrngements should, t lest, over priority determintion for the intr- nd inter-setorl (irrigtion, muniiplities, industry, reretion, fishery et.) llotion of wter, nd proper priing poliy to reover full supply osts of wter from the benefiiries. Turkey is resistnt towrd mking ny rdil hnges in griulturl nd wter mngement poliies. However limte hnge, surging griulturl ommodity pries nd rurlurbn ompetition for wter resoures re expeted to ffet the resoure llotion within griulture. Further these ftors will lso hve signifint implitions for the inter-setorl llotion of resoures. The impt on produers nd onsumers will be diverse, thus it is importnt to onsider the eonomy-wide feed bk effets of the onduted senrios to design nd prioritize proper poliy responses for the griulturl nd wter setors. 2.2 Overview of the Agriulturl Setor in Turkey The verge nnul growth rte of the non-griulturl vlue-dded in Turkey during the pst four dedes ( ) ws 5.1 perent. The growth of the griulturl setor ws limited to 1.3 perent per yer. The growth rtes of griulturl prodution for different dedes re presented in Tble 2.7. Reltively high growth dedes re followed by stgnnt periods. The only exeption ourred in , following the implementtion of the struturl djustment 15

18 progrm. With stble poliies nd fvorble limti onditions the setor grew 2 perent nnully during this period. Prodution delined in 2007 due to n unpreedented drought. Tble 2.7. Agriulturl vlue-dded: growth nd shre, GDP series Period/yer 1987 bse 1998 bse Averge nnul growth in griulture (perent) Shre of griulture in GDP (perent) Soure: lulted from TurkSTAT (2008) Shre of griulture in GDP delined drstilly by roughly 20 to 25 perentge points, due mostly to higher growth in the rest of the eonomy. Reent popultion ensus in 2007 indited tht popultion in Turkey grew by 1.9 perent per yer during the sme period. The struture of the griulturl prodution in Turkey hs mny of the hrteristis of developing eonomy. The shre of griulture in totl employment is still round 27 perent. The verge lnd nd herd size per frm household re smll. Frms in Turkey re generlly fmilyowned, smll, nd frgmented. The verge ultivted re per holding ws bout 5.2 h in 1991, nd it inresed to bout 6 h in About 85 perent of holdings, on 41 perent of the lnd, were smller thn 10 h. 15 perent of holdings were from 10 to 50 h, nd they ultivted lmost hlf of the ultivted lnd (TurkSTAT, 1994, 2004b). The verge size inreses from west towrd southest due to the limte nd fertility differenes. Aording to the griulturl surveys, the proportion of the irrigted lnd in totl ultivted lnd inresed from 14 perent in 16

19 1991, to 20 perent in The shre of irrigted lnd is muh higher in the west thn elsewhere in Turkey. One third of the holdings smller thn 1 h hve ess to irrigtion. The distribution of griulturl lnd remined skewed with Gini oeffiient of A slight tendeny towrds the medium rnges from smller sizes re observed from 1991 to 2001 (TurkSTAT, 1994, 2004b). Of the 26 million h of ultivted lnd (TurkSTAT, 2006), field rops hve oupied slightly over 85 perent of the ultivted re sine The shre of the vegetble re is bout 3 perent, but hs been inresing stedily. Orhrds oupy 10 perent of the ultivted lnd. Lnd left to fllow is bout 5 million hetres. The vlue omposition of the griulturl prodution diverge drstilly from the use of the ultivted lnd. The weight of rop prodution hs been dominnt in the totl vlue of prodution. The vlue of livestok produts mkes qurter of the totl vlue (TurkSTAT, 2006b). The struture of prodution is fr from refleting the poliy weights tht seem to underly government intervention in griulture. The poliies re generlly trgeted towrds erels nd industril rops, wheres vegetbles nd fruits hve reltively smller importne prt from some speilty produts. However, the shre of fruits nd vegetbles in totl vlue is slightly over 40 perent. Protetion nd government support of niml produts hve not been suffiient to ounterblne the dditionl osts of feed due to interventions in the erels. Turkish onsumers end up pying higher pries ompred with the verge met nd milk pries in the EU. The employment retion pity of the eonomy hs lwys been problemti, minly beuse the growth in the ountry s pitl stok hs not been ommensurte with the rpid expnsion of the lbor fore. Despite improvements in eonomi inditors sine 2002, the unemployment rte remins stgnnt t round 10 perent. The rurl unemployment rtes, both mle nd femle, re the mjor ontributing ftors in the stikiness of the overll unemployment rte. The delining trends in the rurl lbor fore prtiiption rtes nd the shre of griulture in rurl employment, ombined with inresing rurl unemployment rtes, signl the strt of mjor trnsformtion in the use of lbor in griulture. However, griulture is still helping to overome the hroni nture of unemployment in Turkey. It eses the detrimentl effet of the lk of humn pitl on growth rtes of the lbor fore, nd the inbility of the non-frm setor to pull even more lbor from griulture. The illitery in griulturl employment is signifintly higher ompred to the rest of the eonomy (Çkmk nd Akder, 2005) whih 17

20 ontributes to the diffiulty of pulling lbor out of griulture. Due to the smll verge frm size, griulturl employment hs reltively lrge shre in totl employment. The setor provides employment for lmost ll femles within rurl res with lmost n 85 perent shre in rurl employment. However, like other rpidly growing eonomies, the shre of griulturl employment in overll employment, s well s bsolute griulturl employment, re stedily delining. Agriulturl employment ws 6 million in 2007 ompred with 9 million in the erly 1990s. Turkey my be onsidered perfet exmple of the mismngement of griulturl poliies prtiulrly fter reforms tht took ple in the mid-1980s. Agriulturl poliies involved minly trnsfers nd were not imed t improving produtivity. The trnsfers to produers ourred mostly from onsumers through support purhses for mjor rops bked by high triffs. 1 Until the onset of the struturl djustment progrm in 2001, trnsfers to frmers from the txpyers were not substntil but were ompnied by huge finnil osts. Most of the budgetry trnsfers to frmers were not plnned using high finnil losses for stte bnks. The finnil burden ws further mplified by the duty losses of stte eonomi enterprises through support purhses nd revolving redit lines to the griulturl sles oopertives unions. The Agriulturl Reform Implementtion Projet (ARIP) begn in 2001 s prt of the seond phse of the struturl djustment progrm. Totl produers subsidy in Turkey showed signifint inrese prior to the strt of struturl djustment progrm in The ontribution of griulturl poliies to the frmers revenue inresed from USD3.4 billion to USD8.0 billion during the 1990s (OECD, 2006). The generl effets of ARIP were signifint with sudden drop in the support to griulture in The stte intervention in the output mrkets ws severely restrited in 2001, oupled with the delyed implementtion of diret inome support. The domesti mrket hs been djusting fst. The mrket prie support provided by the border mesures hs piked up gin in 2002 nd it hs remined high ever sine. The shre of totl griulturl support in GDP ws 6 perent in the lte 1990s. It delined to 3.8 perent in 2005, but is still one of the highest (s shre of GDP) mong OECD member 1 Turkey omplished signifint liberliztion of trde in industril produts. The liberliztion in the griulutrl setor hs been proeeding t slow pe. Exept for the primry ommodities extensively used s intermedite inputs in export oriented mnufturing industries (otton, rw hides nd skins), Turkey hs high levels of protetion in met, diry produts, sugr nd bsi erels. 18

21 ountries. The rte of onsumer subsidy equivlents (CSE) is bk to the pre-risis level in 2005 of 21 perent. The distribution of trnsfers to produers hs not hnged muh sine the 1980s, exept in The shre of mrket prie support in produer subsidy equivlents (PSE) remined round 80 perent. The remining burden flls on the txpyers. Signifint shifts between poliy tools ourred in budgetry support. Input prie intervention lmost disppered, insted re bsed diret inome support (DIS) ontributed 15 perent of support to produers. Conerning the trde poliy, high triff levels for the mjor ommodities nd non-triff protetion styed intt until the reent surge in griulturl pries in 2006 nd Considering the ft tht the shre of food expenditures for the verge onsumer is still more thn 30 perent, the government tried to derese the wedge between domesti nd world pries by grnting duty free imports mostly to the stte prourement geny. The funds used for DIS pyments hve been direted more towrds the ommodity speifi defiieny pyments. The stte of griulture both in terms of its growth pttern nd overll struture of prodution mkes it neessry to evlute the eonomy-wide implitions of surging world pries in Turkey. 3 The Modeling Frmework The pplition of omputble generl equilibrium (CGE) modeling nlysis on wter mngement issues is reltively new in the literture. CGE Modeling hs mde possible the explortion of eonomy-wide effets of wter poliy. CGE models deling with wter issues n be brodly grouped into five tegories ording to their reserh questions. The first group of models dels with the ompetition of different setors or lterntive user groups for wter. Seung et l. (2000) models the welfre effets of using wter in irrigtion or for reretionl purposes. Brind (2004) on the other hnd, introdues drinking wter demnd nd nlyzes the ompetition between drinking nd irrigtion wter. The seond group of models investigtes the ost reovery nd priing bsed wter onservtion poliies. Beritell (2006) for exmple nlyzes the globl nd ntionl level eonomi impts of wter trnsfer projets in Chin. Vlezquez (2007) nlyzes the effets of the inrese in the prie of irrigtion wter on the effiieny of the wter onsumption in griulture nd the possible rellotion of wter to the other setors. Letsolo et l. (2005) test the triple dividend hypothesis to see if wter prie poliies n bring bout redued wter use, 19

22 more rpid eonomi growth nd more equl inome distribution simultneously. He onludes tht it is possible to hieve triple dividends through wter priing. A third group of models is relted to the filittion or liberliztion of irrigtion wter trde. In ft, lmost ll relevnt ppers in the literture n be inluded in this group. However, some studies re missing the neessry onstruts to simulte wter mrkets. Goodmn (2000) shows tht wter trde n reple onstrution of new irrigtion filities by inresing effiieny. Peterson (2004) shows tht the impt of wter shortges n be ompensted by inresing wter trde. Dywer (2005), extends the nlysis of Peterson (2004) to urbn wter usge. Tirdo (2004) shows the effet of hving mrket for wter rights between urbn nd griulturl setors nd rgues tht suh mrket would benefit both user groups. Kohn (2003), on the other hnd, investigtes the effet of interntionl wter trde by using Hekser-Ohlin frmework. The lst group of models ttempt to ombine CGE models with other types of models. Finoff (2004) introdues bio-eonomi model bsed on generl equilibrium pproh while Smjgl et l.(2005) integrtes theoretilly gent-bsed modeling with CGE models. Lstly, some reent models begn to nlyze the miro-mro linkges in wter issues with CGE models. Roe et l. (2005) nd Dio et l. (2005, 2008) use both top-down (trde reform) nd bottom-up (frm wter ssignments nd the possibility of wter trding) linkges. They onluded tht trde reform (top-bottom or mro to miro linkge) hs higher effet ompred to wter reform (bottom-up or miro to mro linkge). Severl CGE models were developed for Turkey imed t nlyzing mroeonomi issues in the 1990s. A seleted list my inlude Hrrison et l. (1993, 1996), Yeldn (1997), Krdg nd Westwy (1999). Strting in the erly 2000s, efforts devoted to develop CGE models for Turkey hve inresed. However, these models re lso gered towrd mro nd trde nlysis with ggregted griulturl setor. CGE models trgeted to nlyze the issues of Turkish griulture re reltively few. Agriulturl CGE models on Turkey generlly seek to ddress trde liberliztion nd reform issues. The first serious ttempt to nlyze the Turkish griulture by using CGE model is Ckmk et l. (1996), where prtil equilibrium model is oupled with CGE model. The CGE model hd n ggregted griulturl setor together with three non-griulturl setors. The dynmis in the griulturl setor were ptured vi the setorl model. Further, the simultions 20

23 relted to the griulturl setor were done vi the setorl model, while the CGE simultions imed to evlute the effets of mroeonomi shoks nd to revel the role of the griulturl setor in the mroeonomi djustment proesses. Dio nd Yeldn (2001) developed generl equilibrium model of the Turkish eonomy with detiled griulturl setor. They hve used n inter-temporl CGE model to nlyze the effets of globl griulturl trde liberliztion. Turkey is one of the regions in the model long with Moroo nd other Middle Estern ountries. Agriulture is disggregted into five subsetors; grin rops, vegetbles nd fruits, sugr, other griulture nd niml produts. Their min onlusion ws in fvor of trde liberliztion. Dogruel et l. (2003) used CGE model to explore the fesible lterntives of griulturl reform nd links between the publi setor fisl blnes, umultion ptterns, dynmi resoure llotion, nd onsumer welfre under medium-long-term horizon. The model onsists of six setors. Agriulture is modeled s n ggregte setor. Lnd is not inluded in the model s prodution ftor. Hene the model is not rel griulturl model though the question of the reserh is griulturl. The intention of the CGE model developed by Çırpıı (2008), is to nlyze wter mngement issues in Turkey. The model hs 9 setors, inorporting 4 ftors of prodution. Agriulture is not disggregted. Wter enters griulturl prodution s lnd nd wter ggregte librted with Leontief funtion. The Cobb-Dougls prodution funtion is used in the non-griulturl prodution. The model lks the regionl nd setorl detil in griulturl prodution. Studies using CGE models to ondut griulturl poliy nlysis inorporte griulture s n ggregte setor both in prodution nd onsumption. Thus the poliy simultions n not tke into ount possible intertions within the griulturl setor. The model used in this study nmely, Turkish Agriulturl Computble Generl Equilibrium Model with Wter (TACOGEM-W), inorportes enrihed tretment of griulturl prodution nd onsumption tivities. TACOGEM-W is the first ttempt tht exlusively models Turkish griulture both in regionl nd setorl detils. The model utilizes the results of frm level model developed for miro region to estimte the shdow prie of irrigtion wter. Hene, the rent reted by the low 21

24 pries for irrigtion wter is inluded in the model, linking the miro nd mro spets of irrigtion wter mngement. 3.1 Struture of the Agriulturl Computble Equilibrium Model for Turkey TACOGEM-W is Wlrsin CGE model tht inludes the behvior of the three min setors in the Turkish eonomy: the prodution tivities, the institutions, nd the foreign setor. In order to study the impt of rnge of poliy simultions, TACOGEM-W further disggregtes the eonomy into 20 griulturl nd 9 non-griulturl tivities. Agriulturl tivities re tegorized by field rops, livestok, fishing nd forestry (lssified s other griulture ). Nongriulturl tivities inlude mining, onsumer mnufturing, food mnufturing, intermedites nd pitl goods, eletriity nd gs, wter, onstrution, privte servies nd government servies. Tble 3.1 provides brek-down of the Turkish GDP in tivities for the yer Aordingly, bout 12 perent of gross output ws devoted to griulture, bout 25 perent to industry, nd the remining 63 perent to servies. Tble 3.1. Gross ntionl produt by kind of tivity, 2003 Produt in urrent pries ISIC, Rev. 2 Vlue (Billion TL) Setor shre (%) Agriulture Agriulture nd livestok Forestry Fishing Industry Mining nd qurrying Mnufturing Eletriity, gs, wter Trde Trnsporttion nd ommunition Finnil institutions Ownership of dwelling Business nd personl servies (Less) Imputed bnk servie hrges Government servies Privte non-profit institutions Import duties GDP (in purhsers vlue) Net ftor inome from rest of the world GNP (in purhsers vlue) Soure: TurkSTAT (2005) 22

25 By exmining the employment struture in the Turkish eonomy s of 2003, we see tht 34 perent of totl employment is still in griulture, generting only 12 perent of totl inome, s implied in the previous tble. Tble 3.2. Employed persons by brnh of eonomi tivity, 2003 (000) Shre (%) Totl Agriulture, forestry, hunting nd fishing Mining nd qurrying Mnufturing Eletriity, gs nd wter Constrution Wholesle nd retil trde, resturnts nd hotels Trnsporttion, ommunition nd storge Finne, insurne, rel estte nd business serv Community, soil nd personl servies Soure: TurkSTAT (2005) Among the griulturl tivities, field rops nd livestok re further disggregted into prodution in four min regions nd one miro region of the ountry, s given in Figure 3.1: West, Est, Centrl, Southestern, nd the miro-region, LSCB (Lower Seyhn-Ceyhn Bsin). Fishing, forestry nd non-griulturl tivities remin t the ntionl level. Geogrphilly, The West region inludes Istnbul, Mrmr, Aegen nd the Mediterrnen regions (exept for Adn nd Osmniye provines), the region Est inludes Northestern Antolin, Mid-estern Antolin nd Estern Blk Se Regions, Centrl region inludes Western Antolin, Mid- Antolin nd the Western Blk Se, nd the LSCB miro region inludes Adn nd Osmniye provines. Figure 3.1 depits the geogrphil distribution of the regions defined in the SAM. Appendix Tble A.1 nd Tble A.2. give list of regions nd the provines inluded in the model. 23

26 Figure 3.1. The five regions defined in the SAM A onsiderbly lrge shre of lmost ll griulturl prodution is onentrted in the western prts of the ountry, bsed on the griulturl prodution vlue dt from 2003 (TurkSTAT, 2003). 51 perent of the prodution vlue of ll rops is produed in the western region, while up to 65 perent of ll prodution vlue generted from fruits nd vegetbles re produed in the western region. Nevertheless, some regions re prominent in speifi rops: for exmple, 46 perent of ll otton is grown in the Southestern Antolin region, while 7 perent of ll otton is produed solely in the Adn-Osmniye region. Furthermore, the Centrl region provides the mjority of sugr beet prodution in Turkey with 63 perent of the totl. 24

27 Tble 3.3.Shre of regionl prodution vlue in ntionl prodution vlue, 2003 (Agriulture only) WEST CENTRAL EAST SE_ANATOLIAN LSCB TOTAL CROPS FIELD CROPS Whet Brley Mize Other grins Pulses Tobo Sugr beet Cotton (rw) Oil seeds Pottoes VEGETABLES FRUITS NUTS OTHER CROPS ANIMAL PROD Milk Cttle met Other big niml met Poultry nd eggs Other niml prod Soure: TurkSTAT (2003) The institutions setor (or, blok) of the eonomy inludes households, the government nd the Wter User Assoitions (WUA). Households re disggregted into rurl nd urbn households. Rurl households re further disggregted mong eh other ording to their geogrphil lotion. Hene in TACOGEM-W, there re five rurl household nd one urbn household type (urbn household defined t the ntionl level). Households ern inome from ftors of prodution (lnd, lbor, pitl), engge in expenditures on vrious ommodities, sve, nd lso py txes to nd reeive trnsfers from the government. The identifition of rurl household types links to the vrition in resoure endowments ommon to griulture. 25

28 The government ollets tx revenue from numerous soures: prodution, sles, households, nd imports. The government uses these funds for purhses of goods nd servies t the respetive mrkets nd mkes trnsfers to households, nd engges in publi svings. Regionl tivities of field rops, in ddition to lnd, lbor nd pitl, use wter s n input in prodution. These tivities lso mke pyments to WUAs in their respetive regions. These wter hrges re inome erned by the Wter User Assoitions. In TACOGEM-W, it is ssumed tht this wter inome olleted in eh region is simply trnsferred bk to the respetive rurl household in lump-sum. In Turkey, s of 2003, 87 perent of ll griulturl lnd is devoted to ultivtion of field rops (Tble 3.4). About 23 perent of ll griulturl lnd is under irrigtion, nd similrly, 22 perent of ll field re sown is irrigted (Tble 3.5). Tble 3.4. Agriulturl lnd use, 2003 (000 Hetre) Shre (%) Agriulturl lnd Cultivted field re Are sown Fllow lnd Are of vegetble grdens Are of vineyrds Are of fruit trees Are of olive trees Soure: TurkSTAT (2005) Tble 3.5. Irrigted versus rinfed lnd use, 2001 (hetres) Totl re Irrigted Not irrigted Totl re 15,322,010 3,505,749 11,816,261 Are sown 12,253,912 2,716,529 9,537,383 Fllow lnd 2,737,560 2,737,560 Vegetble nd flower 371, , ,503 grdens Fruit orhrds nd other 1,757, ,590 1,302,372 permnent rops Soure: TurkSTAT (2004b) For 38 perent of the irrigted re, wells re the most importnt soure of wter (Tble 3.6). The seond most importnt soure of irrigtion is strems, followed by the use of dms. The most ommon method to supply wter to rops is flooding, whih ounts for 84 perent of the re irrigted (Tble 3.7). 26

29 Tble 3.6. Irrigted re by irrigtion soure, 2001 Are (hetre) Shre (%) Totl irrigted lnd 3,505, Well 1,316, Spring 352, Strem 1,003, Lke 67, Artifiil lke 99, Dm 556, Other soures 109, Soure: TurkSTAT (2004b) Tble 3.7. Irrigted re by irrigtion system, 2001 Are (hetre) Shre (%) Totl 3,505, Flooding irrigtion 2,865, Sprinkler irrigtion 582, Drip irrigtion 57, Soure: TurkSTAT (2004b) To onstrut the regionlized SAM for the yer 2003 (plese see Tble A.3 for more ggregted version of the SAM for the yer 2003), we ombine vrious dt from 2003 Agriulturl Struture-Prodution, Prie, Vlue Sttistis (TurkSTAT, 2003); Telli (2004); 1998 Input Output Struture of the Turkish Eonomy (TurkSTAT, 2004); 2003 foreign trde sttistis by Underseretrit of Foreign Trde (UFT Website); TurkSTAT 2002 Household onsumption expenditures nd inome survey sttistis (TurkSTAT Website), ARIP Quntittive Household Survey (QHS) ommissioned by the Tresury nd implemented by the G.G. Consulting et l. (onduted in 2002 nd 2004), nd the Turkish SAM for 2001 developed by GTAP. This regionlized SAM inludes informtion on the regionl level employment of lbor nd pitl s well s intermedite input use by rop, long with wter hrge nd irrigted versus rinfed lnd rents by rop. Employment of pitl, lbor nd dt on intermedite input use re lso entered for tivities t the ntionl level in the SAM. Detiled inome, onsumption nd sving informtion for the five-types of households (urbn, nd four rurl) re inluded in the institutions blok of the SAM. Also inluded in the institutions blok re the dt on government onsumption, net tx reeipts, nd publi svings, long with the WUAs ounts. Finlly, import, export nd triff dt onerning the EU-25 nd the rest of the world onstitute the 27

30 foreign setor blok of the SAM (the foreign trde prtners to Turkey re the 25 Europen Union ountries nd the rest of the world). These detiled dt re used to obtin the prmeters for TACOGEM-W. The lgebri struture of TACOGEM-W is bsed on the CGE model developed by Lofgren, et l. (2002). The entire sequene of mthemtil equtions tht define TACOGEM-W is provided in Appendix B. Min hrteristis of the TACOGEM-W inlude: Prodution tehnology in eh tivity is defined by CES funtion of vlue dded nd ggregte intermedite input use; Vlue dded in eh tivity is given by CES prodution funtion of ftors used (lbor, pitl, irrigted lnd, rinfed lnd nd wter, if pplible); Aggregte domesti output is distributed mong domesti use nd exports (EU nd rest of the world) by Constnt Elstiity of Trnsformtion funtion; Composite output supply (of domesti supply nd imports from the EU nd the rest of the world) is of Armington form; All produers tke ftor nd ommodity pries s given, nd re ll profit mximizers; Urbn nd rurl household types in eh region hve simple onsumption pttern in the sense tht they devote fixed shre of expenditures on eh onsumption item. Eh household type hs different onsumption pttern depending on household inome nd svings. Impliitly in this struture, households re ssumed to minimize expenditures on onsumption, tking s given the prie of eh ommodity. Households derive inome from ftors of prodution depending on the household type. Urbn households ern inome from lbor servies nd pitl rent, while eh rurl household erns inome from servies of lbor nd pitl, s well s lnd rents (irrigted nd rinfed) nd inome from the WUA s vi trnsfers from government; The government lso hs fixed onsumption pttern in the sense tht it devotes fixed shre of expenditures on eh ommodity. The government derives inome from vrious types of txes (import, export, prodution, sles, et.) nd lso sves. The government in this model lso ts s n intermediry between the WUA s nd the rurl households in the sense tht the wter hrges olleted from griulturl produers by the WUA s re then distributed to rurl households in their respetive regions by the government. 28