U.S. Proposal for WTO Hong Kong Ministerial Conference: What s at Stake for Cotton Producers? Mohamadou L. Fadiga, Sam Mohanty, Suwen Pan, Mark Welch

Transcription

1 U.S. Proposal for WTO Hong Kong Mnsteral Conference: What s at Stake for Cotton Producers? Mohamadou L. Fadga, Sam Mohanty, Suwen Pan, Mark Welch Respectvely, Post Doctoral Research Assocate, Assocate Professor, Research Scentst, and Research Assocate, Department of Agrcultural and Appled Economcs, Texas Tech Unversty, Lubbock, Texas. Box 42132, Lubbock, Texas 79409, Emal: Mohamadou.fadga@ttu.edu Selected Paper prepared for presentaton at the Amercan Agrcultural Economcs Assocaton Annual Meetng, Long Beach, Calforna, July 22-26, 2006 Copyrght 2006 by Mohamadou L. Fadga, Sam Mohanty, Suwen Pan, and Marc Welch. All rghts reserved. Readers may make verbatm copes of ths document for non-commercal purposes by any means, provded that ths copyrght notce appears on all such copes.

2 U.S. Proposal for WTO Hong Kong Mnsteral Conference: What s at Stake for Cotton Producers? Abstract Ths study analyzed the cost to U.S. cotton producers of two polcy alternatves under whch the U.S. seeks to cut ts total AMS payments for cotton by 60%. We consdered two scenaros; the U.S. decdes to act unlaterally versus conductng the polcy ntatve along wth multlateral tarff and subsdy elmnatons from the Rest of the World. The study found a 12% cut n target prce and 8% cut n loan rate are necessary to reach the 60% AMS targeted reducton under the unlateral scenaro. In that regards, U.S. net farm ncome decreases consderably despte an apprecaton of U.S. farm prce. Under a multlateral trade lberalzaton from the Rest of the World, a 9% cut n the loan rate and 4% n loan rate are enough to reach the AMS reducton threshold. The study found there s 20% chance that net farm ncome would apprecate and 80% chance that t would declne. However, the declne s less severe compared to the stuaton where the U.S. acts alone. Overall, the sole benefcares n both polces are manly the major exporters such as Brazl, Australa, West Afrca, and Uzbekstan. Key Words: Unted States, Hong Kong, cotton subsdes, tarff, net farm ncome JEL Classfcaton: Q11, Q17 2

3 Introducton The Mnsteral Declaraton that emerged from the recently concluded World Trade Organzaton (WTO) mnsteral conference n Hong Kong encouraged member countres to contnue ther efforts to reform and lberalze the world cotton market (WTO, 2005). The emphass on cotton may ndcate that agreement n ths area may open the door to broader agreement on the agrcultural sector n general. It may also llustrate recognton of the nexus between trade and development and the potental role of cotton as an engne of economc growth for some of the world s least developed countres (LDCs) partcularly n Sub-Saharan Afrca (SSA) and Central Asa. The SSA countres raw fber exports as a share of total producton have ncreased from 60% n 1980/82 to 85% n 2000/02 and as a share of world cotton trade from 6.9% to 17.3%. For the Central Asan countres, however, although cotton producton declned by almost 30% between 1989/90 and 2003/04, total exports as a share of total producton remans above 70%. Whle producton n these two regons seems to be movng n opposte drectons, cotton stll has a vtal place n the ndvdual countres overall economy. Cotton contrbutes between 4% and 10% of the GDP and between 20 and 45% of total export earnngs for Burkna Faso, Chad, Benn, Mal, Uzbekstan, and Tajkstan (Baffes, 2005). These countres are vulnerable to downturns n world cotton prce because any shortfalls n export earnngs would lead to profound economc damages. Producton and export subsdes from developed countres, partcularly the U.S., are at the core of the controversy surroundng the declnng world cotton prce and the need to create a freer and farer nternatonal trade envronment (Makor, 2005). Although Chna, the European Unon, Turkey, and Egypt provde generous subsdes to ther cotton producers [Internatonal Cotton Advsory Commttee (ICAC), 2005], most of the attenton s devoted to the U.S. farm 3

4 polces. The effects of the U.S. farm polces on the world market are magnfed because of the contnuous declne of the U.S. textle ndustry and the subsequent rse n raw fber exports. The shpments of cotton from the U.S. amounted to 14.2 mllon bales n 2004/05, representng 41% of the world cotton exports (U.S. Department of Agrculture, 2006). The effects of subsdes on foregn markets are hard to prove on both economc and nsttutonal grounds. As Cross (2006) stated, causal relatonshps between subsdes and economc prejudce are dffcult to ascertan because of the cross-effects of other factors such as such as foregn textle manufacturng actvtes, ol prce, and polyester prces. Moreover, countres also use the peace clause argument to undermne any challenge of ther subsdy programs or devse other polcy ntatves to reman below ther mandated AMS celngs. Thus, whle provsons n the Farm Securty and Rural Investment (FSRI) Act of 2002 allows producers to update ther acreage and yeld payment bases, nsttutonalzes the counter-cyclcal payments (CCP), and ncreases crop nsurance subsdes whle lowerng the deductble from 35% to 25% were all perceved as ncentves for overproducton (Sumner, 2003), provng ther detrmental effects on trade reman another matter. Currently, the focus s on the CCP and the marketng loan programs, two polcy schemes consdered trade-dstortng and classfed as Amber Box. The polces that fall under ths category were targeted for cuts under the Uruguay Round Agreement on Agrculture. These two polces determne the Aggregate Measure of Support (AMS). The CCP program, whch started n 1998 as an emergency fund to provde ncome support to U.S. producers when world cotton prces fell to hstorcal lows, became permanent under the FSRI Act. The CCP program s lnked to the market prce of a specfc commodty and s trggered whenever effectve market prce falls below the target prce. The overall defcency payments were about 17% of total government outlays n 1998 when t started as an emergency fund to grow up to 25% n 2003/04 under the 4

5 CCP program. The 2002/03 payments were relatvely lower because of hgher nternatonal prces. The other component of the Amber Box, the Loan program, enables producers to hold ther crops when prces are at or below the loan rate. Producers have the opton to use ther crop as collateral to take out loans. Partcpatng producers may choose to repay the loan at a lower repayment rate for marketng loan gans. Producers may also elect to receve marketng loan benefts through loan defcency payments. The loan program s wdely used and payments for marketng loan benefts (.e., marketng loan gans and loan defcency) represent a sgnfcant porton of U.S. farm program outlays from 43% n 1997/98 to 76% n 2001/02. Thus, the enactment of the FSRI Act was perceved as a lack of commtment to free trade on the part of the U.S. (Sumner, 2003; Makor, 2005) because as components of the new farm bll, these programs are expected to reman n place for the next fve years. To address what s generally perceved as a lack of transparency and equty n the world tradng system and address the cotton subsdy problem n a genune manner, the Offce of the U.S. Trade Representatve floated a proposal before the Hong Kong Mnsteral summt. The proposal advocated major reforms n all three areas of concern dentfed n the Doha Development Agenda (.e., domestc support, market access, and export competton) by all member countres, wth some specal and dfferental treatment for developng countres (U.S. Trade Representatve, 2005a). Wth respect to domestc support, the proposal ncluded a 60% reducton n the fnal bound total aggregate measure of support (AMS) for the Unted States (US$19.2 bllon to US$7.6 bllon) and an 83% reducton n the fnal bound total AMS for the European Unon and Japan over a fve-year perod. For all other countres, except those classfed as LDC, the proposed cut was 37% of the total bound AMS level. In the areas of market access and export competton, the proposal ncluded substantal reductons n tarffs wth 5

6 deeper cuts for hgher tarffs and a complete elmnaton of all forms of export subsdes by 2010 for all products. The Hong Kong Mnsteral Declaraton s n prncple smlar to the U.S. proposal. It too advocates an elmnaton of all export subsdes and meanngful reductons of tarffs and domestc support. However, t calls for the establshment of bands of AMS support and bands of tarff protecton by whch members offerng hgher levels of domestc support or hgher tarff rates wll be requred to accept deeper cuts and reductons. The declaraton agreement provdes no specfc thresholds or defntons of these bands and no consensus has been acheved regardng how much each respectve band wll be cut. Notwthstandng the proposal currently on the table, the Afrcan quartet (Benn, Mal, Burkna Faso, and Chad) co-sponsors of the sectoral ntatve on cotton (WTO, 2003) countered wth ther own proposal on March 1, They advocated deeper cuts for cotton compared to agrculture n general, changng the base perod to , and establshng a fund through blateral and multlateral cooperaton to supplement ther ncome shortages through the transton perod leadng toward a world free of subsdes (WTO, 2006). The quartet vewed ther new proposal consstent wth the General Councl August 2004 mandate to address the cotton problem n an ambtous, expedtous, and specfc manner (WTO, 2005). Although a complete elmnaton of export subsdes and substantal cuts n domestc supports and tarff bndngs would be major steps to reformng the world trade system, substantal reforms to mprove agrcultural and nonagrcultural market access from all countres are necessary for the developng countres to realze ther potental gans from agrcultural trade (Anderson and Martn, 2005). Ths argument stems from the fact developng countres are tradng among themselves more than ever before and t would be shortsghted to focus solely on developed countres. 6

7 Whle the offce of the U.S. Trade Representatve ctes sgnfcant support from the Amercan agrculture sector for ts proposal (U.S. Trade Representatve, 2005b), the fundamental queston that arses s how the U.S. proposal to lberalze agrcultural markets would affect U.S. producers. Ths queston may be of specal nterest to the U.S. cotton sector snce the cotton market has been targeted for specal consderaton. Snce the U.S. cotton program s desgned to provde ncome support to cotton producers, what would be the effect of substantal reductons n levels of domestc support on the net farm ncome of U.S. cotton farmers? Would mprovements n market access around the world called for n the proposal offset potental losses n U.S. net farm ncome followng substantal reducton n the AMS? To answer these questons, a partal equlbrum econometrc model of the world fber market, developed by the Cotton Economc Research Insttute (CERI) at Texas Tech Unversty, s used. The analyss consders two scenaros under whch U.S AMS s reduced by 60%. The frst scenaro analyzes the effects of ths proposal on world and domestc cotton prces, U.S. government outlays, gross and export revenues, and net farm ncome f the U.S. proceeds unlaterally. The second scenaro analyzes these same effects as well as effects on cotton mports and exports of selected natons f the U.S. polcy change s accompaned by mult-lateral trade reform (cuts n U.S. prce support and the elmnaton of tarffs and cotton prce supports nternatonally). Conceptual Analyss The conceptual analyss followed n ths study bulds on a prevous analyss of the effects of U.S. cotton programs on the world market (Pan et al., 2005) and a study by Tokarck (2003) on the welfare effects of producton and nput subsdes elmnaton. The novelty pertans to the analyss of the effects of marketng loan beneft ncludng the marketng loan beneft wedge on 7

8 producton decson, the effects of Chnese tarff rate quota, and the effects of producton and nput subsdes used n other major producng and mportng countres such as Chna, Inda, Pakstan, and Turkey. The CCP program s not specfcally addressed n ths secton although t s understood that the provson of the FSRI act that allows producers to update ther yeld base s lkely to have some mpact on producton decson. Fgure 1 analyzes the effects of the U.S. Marketng loan program on the world cotton market. As panel A-1 ndcates, the loan rate plus the marketng loan beneft wedge ( LR+ w ) acts as a prce floor for U.S. cotton producers. The marketng loan beneft wedge ( w ) accounts for the effects of seasonalty on prce and the dfference between posted county prces and natonal prces (Westhoff, Brown, and Hart, 2005). Gardner (2002) evaluated the average marketng loan beneft wedge at 9% above the loan rate; thus, t s an mportant component of U.S. producers plantng decson. As the prce floor to whch producers respond, the loan rate wth the added marketng loan beneft wedge leads to a knked U.S. supply curve, whch becomes nelastc below LR+ w. Snce the loan program does not affect domestc demand, the rse n producton leads to a declnng domestc prce from PE to PD. Hence, producers who partcpate n the loan program would realze a marketng loan beneft equals to LR+ w PD as long as the domestc prce s below the loan rate. In the world market (Panel B-1), ths translates to a knked U.S. excess supply curve (ESus). The ntersecton of the U.S. excess supply curve and the Rest-of-the-world excess demand curve (EDrow) determnes the prce level (P W1 ) under whch the world market clears. If the U.S. reduces ts loan rate, the graphcal analyss shows a reducton n cotton producton and a slght ncrease n domestc cotton prce. In the world market, the excess supply would shft upward, leadng to a reducton of exports and a slght apprecaton n world cotton prce. If the 8

9 loan rate were elmnated, cotton supply response n the Unted States would no longer be knked resultng n an upward shft of the excess supply functon from S to K. Overall, U.S. exports would decrease from (Q S1 -Q D1 ) to (Q S2 -Q D2 ) and world prce would ncrease to P W2. Fgure 2 llustrates the effects of a tarff-rate quota (TRQ) system and smple tarff schedules used to restrct mports. Chna uses the TRQ scheme, whch causes major dstortons n the fber cotton market because of Chna s place as the world largest cotton mporter (Pan et al., 2005). As part of ts commtment to the WTO Chna has establshed a TRQ system for cotton mports n whch the n-quota mport level Q TRQ s set to 890,000 metrc tons wth a tarff of 1% and the out-of-quota tarff was set to 40% n The effects of the TRQ and smple tarff schemes on the world cotton market was analyzed usng the Morath-Sheldon framework (Morath and Sheldon, 1999) n whch we consder the Chnese and the Rest-of the world mporters separately. As Panel B-2 ndcates, wth the TRQ n place, Chnese mporters face a knked world excess supply curve ES TRQ. The excess supply curve s dscontnuous at Q TRQ, the quota level below and beyond whch foregn exporters respond to prce sgnals. The other mporters who are subject to an ad valorem tax face a straght excess supply curve ES ST. The dstortonary effects of these two border polces depend on the poston of the world excess demand curve relatve to the world excess supply curve (Beghn et al., 2001). Consderng the stuaton n whch the TRQ s bndng, Panel B-2 dsplays the world market equlbrum under a TRQ system and the total mports amounts QTRQ. Under a smple tarff structure, the total world mports would be Q ST, larger than Q TRQ. If all forms of tarffs were elmnated, total mports would ncrease to Q CM found at the ntersecton of ED and the world excess supply curve under a no tarff schedule (ES FT ). Thus, the elmnaton of the TRQ system and smple tarff system would lead to an expanson of trade worldwde. 9

10 The analyss of the effect of nput and producton subsdes removal draws from Tockarck (2003), whch was extended to a two-panel dagram llustrated n Fgure 3. Panel A-3 shows a supply and demand equlbrum under whch nput and producton subsdes are provded to producers. At an effectve prce ( 1 ) PS = PD + s + k, the country supples Q S1 quantty of cotton whle consumng Q D1 determned by the domestc prce. The domestc prce s assumed equal to the world prce, whch s determned from the excess supply and demand equlbrum n Panel B-3. If the nput and producton subsdes are dscontnued, the supply curve shfts nward, leadng to smlar movement of the excess supply curve n the world market (Panel B-3). Ths s followed by an ncrease n the world prce from P W1 to P W2, whch s the new compettve prce equlbrum. Under ths new market condton, domestc demand falls to Q D2 and domestc supply falls to Q S2. Overall, world cotton prce apprecates under the new polcy scheme. The Model The study appled the Cotton Economc Research Insttute (CERI) partal equlbrum model for the world fber market to nvestgate the effects of a 60% reducton of the AMS under the scenaro n whch U.S. acts alone and under a scenaro, whch consders a full trade lberalzaton of the world fber market. The CERI world fber model was used to estmate the effects of domestc and trade dstortons n the world cotton market. Ths model ncluded 24 countres and regons that nclude all major cotton exporters and mporters. The model accounts for producton area heterogenety wthn some countres, substtutablty between cotton and competng fbers, and lnkage between raw fber and the textle-manufacturng sector. For a representatve 10

11 country, the model ncludes supply, demand, endng stocks, and market equlbrum condtons for both cotton and man-made fbers. Cotton producton ( PRD ) was modeled usng separate acreage ( ACR ) and yelds ( YLD ) equatons. The acreage response model was specfed as a functon of expected net return of cotton ( ENR ) and competng crops ( ENR ) and a tme c trendt. Smlarly, yeld was also specfed as a functon of expected farm prce ( FPR ) and lag of ranfall ( LRF ). Lastly, cotton producton n country was derved by smply multplyng yeld by total acreage. The full producton model s as follow: β ( ) ( ) α ACR = T ENR ENR c c o ϕ ( ) ( ) φ YLD = T FPR LRF PRD = ACR YLD c c c ρ γ o Fber demand estmaton followed a two-step procedure that connects textle output to fber nputs. The frst step nvolved the estmaton of total domestc textle producton that s total fber demand ( DM f ) from whch the demand for all fbers was derved. In the second step, total domestc textle producton was allocated among the varous fbers manly cotton, manmade fber, and wool. Thus, demand for each fber type was calculated based on ts utlzaton n the textle producton process. The total fber demand ( DM ), total cotton demand ( DM ), and cotton endng stock ( ES ) were specfed as follows: f σ ( ) ( ) f τ DM = DM FPI GDP ( / ) DM = PD PD ρ c c m ( ) ( ) ( ) ω ϖ ψ ES = BS PRD FPR c 11

12 where DM s a constant representng the autonomous consumpton, FPR s the fber prce ndex, GDP s the gross domestc product, PD and c m PD are the domestc prces of cotton and man-made fber, respectvely. Man-made-fber producton was modeled usng estmatons of producton capacty ( CPT ) and capacty utlzaton ( CPU ). Man made fber producton capacty s determned by lag of man-made-fber domestc prce ( LPD ), lag of ol prce ( LPD ), and lag capacty ( LCPT m m ). Total capacty utlzaton s dctated by current domestc man-made-fber prce ( PD ) and current ol prce ( PD ), and lag of capacty utlzaton ( LCPU m l l m ). Total man-madefber producton ( PDR ) s derved by multplyng the producton capacty by the capacty m utlzaton. The full man-made-fber producton model s specfed as ( ) ( ) ( ) η µ κ CPT = LPD LPD LCPT m l m m θ l ( ) ( ) m m κ m CPU = PD PD LCPU PRD = CPT CPU m m m Exports demand ( XPD ) was modeled as a functon of the rato of nternatonal prce of cotton ( P W ) and domestc prce of cotton ( PD ) wth the nternatonal prce n domestc c currency. The mports demand equaton ( IMD ) s a functon of nternatonal prce, exchange rates ( XR ), tarff rates t, and quota restrctons. The domestc market equlbrum s as follows c ( ) XPD = P XR PD ε W ( ( 1 ) ) IMD = P + t XR π W 12

13 ES + DM + XPD = BS + PRD + IMD Solvng ths equlbrum yelds the domestc prce of cotton. Moreover, at the world level, total exports by all countres equal total mports by all countres. The equlbrum n the world cotton market s XPD = IMD. The cotton world prce (A-ndex), domestc cotton prce, cotton textle prce ndex, non-cotton textle prce ndex, farm prce, and man-made fber prce were endogenously solved by equalzng world exports and mports. There are some noteworthy dfferences between countres, especally on the supply sde, namely the specfcaton of the net return equaton. Whle a per acre net return values s derved for the U.S., n most countres, an effectve prce ( PS ) s solely used. As n Tokarck(2003), the effectve prce s derved from the farm prce, the total producton subsdes s, and nput subsdes k. The effectve prce equaton s specfed as follows ( 1 ) PS = FPR + s + k For the U.S., the expected net return s derved as ENR= EFPR YLD TVC+ 0.5 ECCP+ EMLB where EFPR s the expected farm prce, TVC the total per acre varable cost, ECCP s the expected countercyclcal payment rate, and EMLB the expected marketng loan beneft rate. The expected marketng loan beneft rate s the loan rate less the loan repayment rate that s the lower of the loan rate or the adjusted world prce; t s derved as ( mn (, ) ) EMLB= LR LR AWP + w LYLD 13

14 where LR s the loan rate, AWP s the adjusted world prce, LYLD s the lag of yeld, and w remans as prevously defned. The dfference between the target prce and the effectve market prce (.e., drect payment rate plus the hgher between the loan rate and the average market prce) represents the CCP rate. The expected countercyclcal payment s derved as ( max (, )) ECCP= TP DP+ LR LPD BYLD where TP s the cotton target prce set at 72.4 cents/lb. under the 2002 FSRI Act, DP the drect payment set at 6.67 cents/lb., LR s the loan rate set at 52 cents/lb, and BYLD s the base yeld, whch corresponds to the average yelds. In the above specfcatons the superscpts n greek letters are the coeffcents to be estmated usng hstorcal data collected from varous sources. The models were lnearzed by log transformatons. Thus, the coefcents are the elastctes and may also be nterpreted as mpact multplers and determne the magntudes of the smulaton results. Detaled results of the estmaton and the derved elastcty estmates are avalable n Pan, et al.(2004) or from the authors upon request. The data used n the study were compled from varous sources. The hstorcal and predcted macroeconomc varables (real GDP, exchange rate, populaton, and GDP deflator) were obtaned from the Food and Agrcultural Polcy Research Insttute (FAPRI). Cotton producton, consumpton, endng stocks, mports, and export data were retreved from U.S. Department of Agrculture Producton, Supply & Dstrbuton (PSD). Fber mll consumpton and man-made fber data were retreved from the Food and Agrculture Organzaton of the Unted Natons (FAO) World Fber Consumpton Survey (before 1994) and Fber Organon (after 1994). 14

15 Polcy Shock and Assumptons The approach was to develop a fve-year baselne (2006/ /11) assumng a contnuaton of current domestc and border protecton polces. Then a 60% reducton of AMS was targeted fve years after the begnnng of mplementaton holdng all other polces unchanged. The threshold of 60% reducton was reached through lnear cuts of the target prce by 12% and the loan rate by 8% usng a progressve formula of equal ncrements. The world cotton market was then allowed to react to the resultng prce sgnals over a fve-year perod that s through 20010/11. The effects were measured by comparng the AMS under the baselne to ts value after the polcy changes. Addtonally, the effects of these polces on world prce, U.S. farm prce, producton, exports, government outlays, and net farm ncome were determned by evaluatng ther devatons wth respect to ther baselne values. A second scenaro wth full trade lberalzaton from the Rest of the World establshed was consdered. The full trade lberalzaton was establshed over a fve-year perod followng ncremental lnear cuts of tarff rates and subsdes. In ths case, the 60% AMS reducton threshold was accomplshed by a 9% reducton n the target prce and 4% reducton n the loan rate followng the same formula. The effects of these polces were evaluated on the U.S. sde and the Rest of the World. Fnally, a stochastc analyss of the effects of polcy changes on the U.S. net farm ncome was conducted to generate confdence bands and the cumulatve dstrbuton functon of the changes wth ther assocated probablty levels. The stochastc smulaton follows the same approach as n Fadga, Mohanty, and Pan (2005). Ths approach was based on a multvarate emprcal dstrbuton (Rchardson, Klose, and Gray, 2000) of hstorcal exogenous data to generate 500 correlated random draws of the exogenous varables, 15

16 whch are then substtuted nto the partal equlbrum model to solve for the 500 set of endogenous varables, ncludng U.S. net farm ncome. The number of draws was set to 500 to be consstent wth FAPRI approach to stochastc analyss of agrcultural commodtes (FAPRI, 2004). Moreover, there s no lmt on the number of exogenous stochastc varables to use to generate the draws; however, one has to be mndful of the computatonal cost assocated wth large matrx. In ths study, t was assumed that yelds are at the bass of most of the uncertanty n the world fber market and were used to generate the draws. Smulaton Results The results n Table 1 summarze the effects of a 60% unlateral reducton n AMS on U.S. prces, polcy nstruments, and government outlays. Whle such an acton by the U.S. yelded modest effects on world cotton prces (3.47% on average), the effects on U.S. farm prces were relatvely hgh (5.04% on average). The magntude of the ncrease s due to a contracton of acreage because of reducton of target prce by 12% and loan rate by 8% over 5 years. Loan defcency payments decreased by 40.93% whle countercyclcal payments fell by 100% by 2010/11. A 60.56% reducton of AMS relatve to the WTO base was acheved by 2010/11. Although world prce slghtly ncreased, the results n Table 2 show U.S. cotton export dropped by 4% on average because of producton reducton. Snce the LDP s coupled whle the CCP s assumed 50% coupled n the CERI model, a reducton n these polcy nstruments lead to less acreage, hence to reduced producton level. Ths further explans the declne n the varable cost of producton. Whle gross revenue ncreases because of apprecatng U.S farm prce, the U.S. net farm ncome fell consderably because of substantal reducton n government payments. 16

17 The fall n net farm ncome starts at the begnnng of the polcy shocks and rapdly declnes by 26.01% n 2010/11. Overall, the U.S. treasury was the sole benefcary of a unlateral move from the U.S wth an average reducton n total government payments by 30.82%. As Table 3 llustrates, f the changes n U.S. polcy were conducted wth multlateral trade lberalzaton of the world cotton market, a reducton of the target prce by 9% and the loan rate by 4% were suffcent to acheve the 60% AMS reducton goal. Ths was possble because the U.S. farm prce and the world prce of cotton apprecated more under the multlateral lberalzaton scenaro. A 60% AMS reducton under multlateral lberalzaton nduced relatvely hgh changes n the A-ndex, averagng 10.57% over the smulaton perod. The dynamcs of the changes were also noteworthy, ncreasng steadly as the cumulatve effects of the lnear cuts n the target prce and loan rate ncreases to reach a hgh at 13.28% n 2009/10. In 2010/11, the A-ndex adjusts to new market equlbrum wth a 12.74% change relatve to the baselne. The U.S. farm prce follows a smlar pattern, but the changes were slghtly hgher because of further cuts n U.S. acreage. As n the unlateral polcy mplementaton scenaro, hgher domestc and nternatonal cotton prces under multlateral reform lead to a rapd reducton n LDP and CCP wth the latter declnng by almost 100% n 2010/11. However, these effects were acheved wth much smaller cuts n the loan rate and the target prce than the unlateral scenaro. The results n Table 4 ndcate that U.S. cotton producton and exports declne by 0.55% and 0.66% on average. Projected U.S. net farm ncome values stll lay below ther baselne levels n the multlateral trade lberalzaton scenaro. From a small decrease of 1.76% at the begnnng of the smulaton perod, losses n net farm ncome grew rapdly, reachng 18.72% n 17

18 2009/10. Thus, by these estmates, multlateral lberalzaton only partally offsets the losses n ncome due to the cuts n AMS. For the Rest of the World, the results are analyzed by lookng at the effects of the U.S. trade proposal when accompaned wth multlateral tarff and subsdy elmnatons. For cotton mportng countres (Table 5), the overall effects depend on the degree of protecton that exsted pror to the trade lberalzaton. In Chna for nstance, the elmnaton of the tarff rate quota system and producton subsdes leads to hgher mport demand. As Table 5 ndcates, on average, Chnese mports are, on average, expected to ncrease by 6.72% (over 800 thousands bales) relatve to the baselne. For Pakstan, mports ncrease by over 4.20% on average. Imports by Inda follow a relatvely stable pattern ncreasng by about 4.17% throughout the perod. Turkey and the European Unon provde subsdes to ther producers, whch f phased out would lead to a declne n producton, especally n the case of the European Unon where subsdes are much hgher. The contracton n producton leads to hgher mports, averagng 3.36% above the baselne. Hgher nternatonal prces of cotton are not favorable to Korean, Tawanese, and Japanese textle ndustres, whch rely almost exclusvely on mports for ther operatons. For these countres, mports are expected to declne relatve to ther baselne values. Of the major cotton mporters reported here, the smallest effects from the U.S. proposal are seen n the area of Mexcan cotton mports. Whle Mexcan mports and exports to the U.S. are traded n a free market envronment due to the North Amercan Free Trade Agreement (NAFTA), the removal of cotton mport tarffs n Mexco wll spur a small ncrease n demand wth other tradng partners. As for cotton exporters (Table 6), t was expected that the non- and low-subsdzng countres would capture producton dsplacement from subsdzng and less cost compettve 18

19 countres. Brazl s expected to ncrease ts cotton exports by an average 10.65%, followed by Australa (5.80%), West Afrca (5.49%), and Uzbekstan (4.76%). For Brazl, these export levels were expected n the context of lberalzaton as more area n the expandng fronter regon enters producton because of hgher prces. As for Australa, cotton farmng s a trbutary of water avalablty, whch serves as a constrant on cotton producton and thereby cotton exports. For West Afrca, lmted technologcal nnovaton and contnued subjugaton to weather varablty prevent these countres from takng full advantage of hgher prces. Fnally, fgure 4 captures the probablstc outcome of the changes n U.S. net farm ncome under unlateral and multlateral settngs n 2010/11, the year when the 60% AMS reducton and the trade lberalzaton are fully mplemented. It s mportant to note the dfference n the results obtaned under the determnstc analyss compared to the stochastc analyss. Concernng the unlateral scenaro, the change n net farm ncome under the determnstc analyss (Table 4) amounted to % whle the stochastc mean amounted to %. Under the multlateral scenaro, the determnstc and stochastc means are % and -6.36%, respectvely. There s a strong devaton n the results obtaned under the two modelng strateges. The results show the probablty of an ncrease n net farm ncome under the frst scenaro s close to zero. Moreover, t s hghly lkely (60% chance) that the change n net farm ncome would fall by more than 15%. However, under a multlateral lberalzaton, there would be 20% chance that the net farm ncome would ncrease between 0% and 9.5% whle the lkelhood that the net farm ncome would fall between 5% and 15% s relatvely hgh (60% chance). The average changes n net farm ncome are respectvely % and -6.36% under the unlateral and multlateral scenaros. 19

20 Conclusons Ths study analyzed the cost to U.S. cotton producers of a 60% AMS reducton f the U.S. decded to act unlaterally versus f the polcy ntatve s accompaned wth multlateral trade lberalzaton from the Rest of the World. The study shows that f the U.S. acts alone, substantal cuts n the target prce and the loan rate are needed to meet the targeted AMS reducton (12% and 8%, respectvely). U.S. net farm ncome decreases consderably because of consderable cutbacks n government payments that cannot be compensated by the moderate ncrease n U.S. farm prce. If the polcy change s conducted wth multlateral trade lberalzaton from the Rest of the World, the negatve effects on U.S. net farm ncome are somewhat mtgated, but do not fully compensate for the losses n government prce support (9% target prce and 4% loan rate reducton). Thus, net farm ncome decreases relatve to the baselne n both scenaros. For competng cotton exporters, substantal ncreases n cotton exports from Brazl ndcate that that naton s a prmary benefcary of the U.S. proposal, followed by other leadng cotton exporters Australa, West Afrca, and Uzbekstan. 20

21 References Anderson, K. and W. Martn Agrculture: The Key to Success of the Doha Round n Trade, Doha, and Development: A Wndow nto the Issues, eds. R. Newfarmer, pp Washngton DC: World Bank, Baffes, J. Cotton and the Developng Countres: Implcatons for Development n Trade, Doha, and Development: A Wndow nto the Issues, eds. R. Newfarmer, pp Washngton DC: World Bank, Beghn, J., B. Osta, J. Cherlow, and S. Mohanty. The Cost of the U.S. Sugar Program Revsted. Contemporary Economc Polcy, 21(2003): Cross, K. H. Kng Cotton, Developng Countres, and The Peace Clause: The WTO s U.S. Subsdes Decson. Journal of Internatonal Economc Law, 9, 1 (2006): Fadga, M. L., S. Mohanty, and S. Pan. The Impacts of the U.S. Cotton Program on the West and Central Afrcan Countres Cotton Export Earnngs. Agrcultural Economcs Revew, 2 (August 2005): Gardner, B. "U.S. Agrcultural Polces snce 1995, wth a Focus on Market Effects n Grans and Olseeds," Workng Paper 02-17, Department of Agrcultural and Resource Economcs, Unversty of Maryland, Internatonal Cotton Advsory Commttee (ICAC). Producton and Trade Polces Affectng the Cotton Industry, presented at 64th Plenary Meetng Lverpool, Unted Kngdom on September Makor Chrstne M. Reformng the Cotton Trade Order? An Analyss of Cotton Subsdes and Implcatons for Sustanable Development. Journal of Publc and Internatonal Affars, Vol. 16, sprng 2005, pp Morath T. and I. M. Sheldon, The Economcs of Modelng Tarff-Rate Quotas, USDA/WTO Modelng Group Poston Paper, Pennsylvana State Unversty, February Pan, S., M. Welch, S. Mohanty, M. L. Fadga, and D. Ethrdge. Chnese Tarff Rate Quota vs. U.S. Subsdes: What Affects the World Cotton Market More? Journal of Internatonal Law and Trade Polcy, 6, 2 (2005): Pan, S., M. Fadga, S. Mohanty, and D. Ethrdge. Structural Models of the U.S. and Rest-of-the- World Natural Fber Markets. Workng Paper CER 04-03, Department of Agrcultural and Appled Economcs, Texas Tech Unversty, May Rchardson J.W., S. L. Klose, and A. W. Gray. An Appled Procedure for Estmatng and Smulatng Multvarate Emprcal Probablty Dstrbutons n Farm-Level Rsk Assessment and 21

22 Polcy Analyss. Management. Journal of Agrcultural and Appled Economcs, 32 (August 2000): Sumner Danel Implcatons of U.S. Farm Bll of 2002 for Agrcultural Trade and Trade Negotatons. The Australan Journal of Agrcultural and Resource Economcs, 46(3), 2003, pp U.S. Department of Agrculture. (2006). Economc Research Servce. Avalable onlne at Accessed March 02, U.S. Trade Representatve (USTR). (2005a) U.S. Proposal for Bold Reform n Global Agrculture Trade. October 2005a. Avalable onlne at (2005b). Amercan Agrculture Supports U.S. WTO Doha Agrculture Proposal for Bold Reform, 21 October 2005b. Avalable onlne at World Trade Organzaton (2006). Proposed Modaltes for Cotton Under the Mandate of the Hong Kong Mnsteral Decson. Communcaton of Co-sponsors of the Sectoral Intatve for Cotton. Released March 1, Document Code TN/AG/SCC/GEN/4. (2005). Doha Work Program. Draft Mnsteral Declaraton. Released December 18, Document Code TN/MIN (05)/W/3/Rev2. (2003). Poverty reducton: Sectoral ntatve n favor of cotton. Jont Proposal Presented by Benn, Burkna Faso, Mal, and Chad presented at the Ffth Sesson Mnsteral Conference n Cancun, Mexco, September

23 Table 1. U.S. Proposal n a Unlateral Reform Settng: Impacts on Cotton Prces and Government Programs Payments 2006/ / / / /11 Average (Cents/pound) A-ndex Baselne Unlateral Reforms % Change 2.91% 3.74% 3.71% 3.81% 3.21% 3.47% U.S. Farm Prce Baselne Unlateral Reforms % Change 2.97% 4.30% 5.09% 6.66% 6.19% 5.04% ($US mllon) LDP Baselne Unlateral Reforms % Change % % % % % % CCP Baselne Unlateral Reforms % Change % % % % % % AMS Base Baselne Unlateral Reforms % Change -1.55% % % % % % Government Outlays Baselne Unlateral Reforms % Change % % % % % % 23

24 Table 2. U.S. Proposal n a Unlateral Reform Settng: Impacts on the U.S. Cotton Industry 2006/ / / / /11 Average (Thousand of bales) Producton Baselne Unlateral Reforms % Change -1.84% -2.73% -3.33% -4.04% -4.10% -3.21% Exports Baselne Unlateral Reforms % Change -2.31% -3.53% -4.25% -4.97% -5.17% -4.04% Mll-use Baselne Unlateral Reforms % Change 0.00% -0.02% -0.05% -0.13% -0.19% -0.08% ($US mllon) Net Farm Income Baselne Unlateral Reforms % Change -6.87% % % % % % 24

25 Table 3. U.S. Proposal n a Multlateral Reform Settng: Impacts on Cotton Prces and Government Payments 2006/ / / / /11 Average (cents/pound) A-ndex Baselne Multlateral Reforms % Change 6.47% 8.92% 11.46% 13.28% 12.74% 10.57% U.S. Farm Prce Baselne Multlateral Reforms % Change 5.41% 7.93% 8.85% 10.58% 10.00% 8.55% ($US mllon) LDP Baselne Multlateral Reforms % Change % % % % % % CCP Baselne Multlateral Reforms % Change % % % % % % AMS Base Baselne Multlateral Reforms % Change 1.02% % % % % % Government Outlays Baselne Multlateral Reforms % Change % % % % % % 25

26 Table 4. U.S. Proposal n a Multlateral Reform Settng: Impacts on U.S. Cotton Industry 2006/ / / / /11 Average (Thousand bales) Producton Baselne Multlateral Reforms % Change 0.24% -0.02% -0.45% -1.01% -1.51% -0.55% Exports Baselne Multlateral Reforms % Change 0.32% -0.30% -0.60% -0.96% -1.75% -0.66% Mll-use Baselne Multlateral Reforms % Change 1.00% 1.67% 0.56% -0.27% -0.39% 0.51% ($US mllon) Net Farm Income Baselne Multlateral Reforms % Change -1.76% -6.89% % % % % 26

27 Table 5. U.S. Proposal n a Multlateral Reform Settng: Impacts on Major Cotton Importers 2006/ / / / /11 Average (Thousand Bales) Chna Baselne Multlateral Reforms % Change 5.82% 6.10% 6.96% 7.30% 7.42% 6.72% Inda Baselne Multlateral Reforms % Change 4.52% 4.29% 3.92% 4.01% 4.10% 4.17% Pakstan Baselne Multlateral Reforms % Change 3.42% 3.14% 4.13% 4.76% 5.55% 4.20% Japan Baselne Multlateral Reforms % Change -0.36% -2.25% -2.15% -2.53% -2.41% -1.94% South Korea Baselne Multlateral Reforms % Change -0.60% -1.18% -2.00% -2.92% -3.81% -2.10% 27

28 Table 5. (Contnued) U.S. Proposal n a Multlateral Reform Settng: Impacts on Major Cotton Importers 2006/ / / / /11 Average (Thousand bales) Tawan Baselne Multlateral Reforms % Change 0.01% -3.58% -4.06% -4.01% -4.39% -3.21% Mexco Baselne Multlateral Reforms % Change -0.12% 0.48% 0.95% 1.32% 1.55% 0.83% European Unon Baselne Multlateral Reforms % Change 3.20% 3.18% 3.63% 3.24% 3.56% 3.36% Turkey Baselne Multlateral Reforms % Change 2.95% 3.19% 3.40% 3.40% 3.16% 3.22% 28

29 Table 6. U.S. Proposal n a Multlateral Reform Settng: Impacts on Major Cotton Exporters 2006/ / / / /11 Average (Thousand bales) Australa Baselne Multlateral Reforms % Change 4.39% 5.05% 5.36% 6.36% 7.86% 5.80% Brazl Baselne Multlateral Reforms % Change 7.87% 9.55% 11.21% 11.69% 12.91% 10.65% Uzbekstan Baselne Multlateral Reforms % Change 4.24% 4.82% 5.23% 4.65% 4.87% 4.76% West Afrca Baselne Multlateral Reforms % Change 4.68% 5.09% 5.54% 5.94% 6.18% 5.49% 29

30 Fgure 1. Effects of Removng U.S. Loan Program on World Market LR+w D S ES US Pw 2 Pw 1 PE K ED ROW PD S Q D2 Q D1 Q S2 Q S1 Q X2 Q X1 Panel A-1 Panel B-1 30

31 Fgure 2. Effect of Removng TRQ and Smple Tarff on World Market ES TRQ P A P W + t P W ES ST ES CM ED S D Q S1 Q S2 Q D1 Q D2 Q TRQ Q ST Q CM Panel A-2 Panel B-2 31

32 Fgure 3. Effects of Removng Producton and Input Subsdes on World Market PS 1 PS 1 -κ κ S 1 ES 2 ES 1 P D2 Pw 2 P D1 Pw 1 ED S 2 D Q D2 Q D1 Q S2 Q S1 Q X2 Q X1 Panel C-1 Panel C-2 32

33 Fgure 4 Cumulatve Dstrbuton of Change n U.S. Net Farm Income under a Unlateral vs. Multlateral Acton n 2010/11 1 Probablty % -15% -10% -5% 0% 5% 10% 0.2 Multlateral Unlateral 33