Net Versity

ExecutiveSummary

Global food prices nearly doubled during 200408 and have remained relativelyhighsincethen.Mostrecently,theFAOindexofrealglobalfood pricesrosefrom151pointsinJune2009to172pointsinJanuary2010.Therise in global food prices was highest for cereals, which remain relatively expensive:between2005and2008theinternationalpriceofwheatmorethan doubled,andglobalriceandmaizepricestripled,andasofJune2009,wheat andmaizepricesremainedsubstantiallyhigherthanfouryearspreviously(by respectively55percentand87percent)whilericepriceswereaboutdouble.

Anumberofsimultaneouseventsexplaintheunusuallyhighfoodprice inflation that took place during 200708. While supply constraints ( in particular low levels of world cereal stocks) played a role, the main drivers were increases in demand (especially the rapid increase in the use of food cropstoproducebiofuels),speculation(largeflowsofspeculativecapitalinto agriculturalcommodityfuturesmarkets)andpolicyfailures(especiallyexport restrictions).

InSouthAsia,foodpriceinflationvariedsignificantlyamongcountries. In200708,itrangedfromrelativelymoderateinIndia(about7percent),to high in Nepal and Bangladesh (about 15 percent), to very high in Pakistan ( around20percent)andinSriLankaandAfghanistan(morethan30percent). Besidestheintercountryvariations,thereweresignificantvariationsamong commoditiesand,inmanycountries,amongregions.

Food price inflation exceeded nonfood price inflation throughout South Asia except in India. During the 200708 food crisis, food price inflation became the main driver of general inflation throughout of most SouthAsia.Thoughinthesecondhalfof2008,theroleoffoodpriceincreases diminished in most countries of the region after prices came down, in 2010  foodpricesagainbecamethemainfactordrivinggeneralinflationinanumber ofcountriesincludingIndia.

South Asian countries responded to food price inflation with a wide range of policies and measures.Attheheartoffoodpricepoliciesinthe region is the political economy of the tradeoff between consumers’ and producers’interests.Mostofthemeasuresthatweretakensoughttodampen the negative effects of the food price rises on consumer welfare, and were shortterm in nature; measures to address traditional constraints on agriculturalproduction,ortootherwisefacilitateasupplyresponse,havebeen muchlesscommon.Mostpopularhavebeentradepolicymeasurestodampen priceincreases(e.g.abolishingimporttariffs)ortoensureadequatesupplies ondomesticmarkets(e.g.restrictingexports);buildinguporexpandingpublic grain reserves; controlling prices; and extending existing social protection measures or (to a much lesser extent) introducing new social protection programs.

IMPACTOFFOODPRICEINFLATIONONTHEPOOR

Besides negatively affecting macroeconomic stability, food price inflation decreasesthewelfareofhouseholdswhoarenetbuyers,ratherthansellers,of food. In particular, it threatens the welfare of poorer households, for whom foodtakesalargeshareoftotalspending.

InSouthAsiamosthouseholds,includingthoselivinginruralareas,are netbuyersoffoodandlikelytosufferwelfarelossesfromincreasesin foodprices.Asiswellknown,thevastmajorityofurbanhouseholdsarenet food buyers. But contrary to popular perception, 70  80 percent of rural households in the region are also net buyers of the main grain staples ( rice and/orwheat) .

FortheaveragehouseholdinSouthAsia,foodtakesclosetohalfoftotal spending, compared to only 17 percent in the United States. This high percentage makes South Asian populations very vulnerable to food price increases.

Poor people are likely to have been especially hardhit by high food prices.First,poorpeoplespendalargerproportionoftheirincomeonfood. Second,thefoodpriceinflationof200708wasespeciallystarkforcereals,and theproportionofcerealsintotalfoodspendingismuchhigherforthepoor thanthenonpoor.Moreover,householdswhopreviouslywerelivingnotfar above the poverty line are likely to have fallen into poverty as the result of higherfoodprices.

Theeffectoffoodpriceinflationonhouseholdwelfareisanalyzedat twolevelsinChapter2.Thefirst,simpler,levelkeepsquantitiesfixedandis limited to the pure firstorder impact of the price change. Households are classifiedintonetbuyersandnetsellersofacommoditywherethelattergain andtheformersufferwelfarelossesasaresultofapriceincrease.Asecond levelofanalysistakesaccountoftheconsumptionandproductiondecisions that take place as a result of the price change. We use nationally representative household survey data for Bangladesh, Pakistan, and Nepal frombeforethefoodcrisistosimulatetheimpactoffoodpriceinflationon poverty headcount levels, taking account of both firstround and second roundeffects.Theanalysisdoesnotallowforgeneralequilibriumeffects.

In Bangladesh, Pakistan, and Nepal, the net marketing position of householdsleadstosignificantfirstroundwelfarelossesforthelarge majority of households. In Bangladesh and Nepal respectively, about 80  percentand70percentofhouseholdsarenetbuyersofrice,whileinPakistan 77percentofhouseholdsarenetbuyersofwheat.Basedonfirstroundeffects only(netbuyernetselleranalysis)a50percentincreaseinthepriceofrice would raise the national poverty headcount ratio in Bangladesh by about 6  percentage points. Similarly, a 40 percent increase in the wheat price in Pakistanwouldcausea2percentagepointincreaseinnationalpoverty.The estimatedpovertyeffectinPakistanissmallerthaninBangladeshbecauseof the lower price increase and the smaller average gap between household consumptionandproduction.InNepal,a20percentincreaseinthericeprice would raise the national poverty headcount ratio by only 0.5 percentage points, but rising rice prices make households who are already poor even poorer. In all three countries the impact of food price inflation on poverty differssignificantlyamongregions,andregionalpovertyimpactsoftengreatly exceedtheaveragenationalimpact.

In all three countries, secondround responses in production and consumption are found to offset part of the welfare loss from first round effects. In Bangladesh, based on differences in national poverty headcount ratios with and without secondround effects, adjustments in behaviorbyconsumersandproducersdecreasethefirstorderpovertyimpact by about 2 percentage points or up to 30 percent in proportional terms. In Pakistan,thesecondroundresponseshaveanevenlargereffect;theyreverse upto90percentofthefirstroundimpactonthepovertyheadcount.

InSriLanka,foodpriceinflationislikelytohaveincreasedthepoverty headcountratioandtohavebeenparticularlyharmfultothepoorestof the poor. Analysis by the World Bank shows that a large share of the populationwasclusteredaroundthepovertylineevenbeforethefoodcrisis, andimpliesthata10percentdeclineinpercapitaconsumptionmayleadtoa 6 percentage point increase in the poverty headcount ratio. In Afghanistan, thereisnoinformationregardingthepovertyimpactoffoodpriceinflation, butbecauseevenbeforethefoodcrisis42percentoftheAfghanpopulation wasclassifiedaspoor,andanestimated20percentofthepopulationlivedjust above the poverty line, there is no doubt that the impact of the food price crisisinAfghanistanhasbeenextremelyserious.InIndia,existingprograms andpoliciesmadethatcountryrelativelyshockprooftothefoodcrisis.

The firstround welfare loss caused by higher food prices is larger in urban areas than in rural. Unlike most urban consumers who can only respond on the consumption side, most rural households can adjust both consumptionandproduction.InBangladeshtheheadcountratiobasedonthe upper poverty line increases by about 6 percentage points in urban areas compared to 5 percentage points in rural areas. In Pakistan the urbanrural disparityisevenlarger:theurbanheadcountratioincreasesby3percentage points while the rural ratio increases by less than one percentage point. In Nepalthedisparitybetweenruralandurbanareasislesspronounced,because ofthesmalleroverallimpactofhigherfoodprices.

Theimpactoffoodpriceincreasesonthepovertyheadcountalsovaries considerablyamonghouseholdsindifferentincomegroups.Bangladesh andPakistanbothhavehighconcentrationsofhouseholdsaroundthepoverty line.Asaresultmostofthewelfarelossfromhighfoodpricesisconcentrated amonghouseholdsinthemiddleofthedistributionofpercapitaexpenditures. In Bangladesh the firstround impact of higher food prices on the poverty headcountratiointhethirdexpenditurequintileis24percentagepointsusing the country’s upper poverty line and 34 percentage points using the lower povertyline.Similarly,inPakistanthepovertyheadcountratiointhesecond (nexttopoorest)quintileincreasesby11percentagepoints.Inbothcountries, higherfoodpricesleadtoaslightdecreaseinthepovertyheadcountinthe poorestexpenditurequintile.Thisisbecauseeveninthepoorestgroupsthere aresomehouseholdsthatarenetsellersandthereforegainfromthefoodprice increase.Sincebeforethefoodpriceriseallhouseholdsinthepoorestquintile inPakistanwerepoor,theheadcountratiodecreasesassoonasoneormoreof thesehouseholdscrossesthepovertyline.

Households that remain below the poverty line are worse off with higher food prices.InBangladeshthesimulatedricepriceriseleadstoan increase in the intensity of poverty as measured by the poverty gap which increasesfrom0.35to0.41(upperpovertyline)and0.25to0.32(lowerpoverty line).Similarly,inPakistanthesimulatedwheatpriceincreaseresultsinarise in the poverty gap from 0.17 to 0.18. Thus, high food prices clearly hurt the pooresthouseholds.

Despitethepossibilityofsecondroundeffects,aconsiderableportion ofthewelfarelosscausedbyfoodpriceinflationislikelytopersist— unless the general equilibrium effects are significantly larger than second roundadjustments.Thatoutcomeisunlikely,giventhatthewageelasticities of output prices in South Asia tend to be relatively low. Whether poverty causedbyincreasingfoodpricesispermanentortransitorywilllargelydepend on whether high food prices persist and for how long, and on how governmentsrespond,intermsofsocialprotectionprogramsandotherpolicy measures.

TRADELIBERALIZATIONANDFOODPRICES

In principle, intraregional trade liberalization could mitigate food price inflation. Discussions regarding the food crisis in South Asia have largely ignored the regional dimension of food price inflation and the possibilityofimprovingfoodsecuritybyliberalizingtrade.Incountriesthat traditionallyrelyonfoodimports,regionaltradeliberalizationmightincrease confidenceininternationalmarkets.

During the food crisis most countries in South Asia increased their trade barriers instead of facilitating trade. Whileinanefforttocontrol domestic food prices, most South Asian countries reduced import taxes, several of them also introduced export control measures or even banned exports of certain staples. These “beggarthyneighbor” type policies aggravated price increases elsewhere, as seen in Afghanistan where wheat prices shot up after Pakistan introduced an export ban, and in Bangladesh where India’s restrictions on rice exports contributed to rice price inflation. Exportbansalsoencouragedsmugglingwhileloweringeconomicreturnsfor domesticfarmers.

TheSouthAsianFreeTradeArea(SAFTA)agreementaimsatincreasing intraregionaltradeviapartialtradeliberalization.Basedonformaltrade flows, South Asia is one of the world’s least integrated regions. The SAFTA agreementisanattempttoincreaseintraregionaltradethroughthegradual dismantling of some tariff barriers, but it leaves out a large number of products denominated as sensitive, and it does not address nontariff trade barriers. Chapter 3 of this report uses a worldwide recursive dynamic computablegeneralequilibriumtrademodeltoanalyzeSAFTA’spotentialfor increasing intraregional trade and mitigating food price increases in South Asia.

ThefindingsshowthatSAFTA’spotentialforinfluencingdomesticfood pricesinSouthAsiaislimited.Themodelsimulationsindicatethatglobal restrictionsoncerealexportshadamuchsmallerimpactondomesticpricesin SouthAsiathantheglobalaverage,mainlybecauseofSouthAsia’srelatively limiteddependenceoninternationalmarkets.TheyalsosuggestthatSAFTA hardlydampensdomesticpriceincreases,mainlybecauseofthelargenumber of“sensitiveproducts”(negativelist)andtheabsenceofagreementsregarding nontarifftradebarriersandsubsidiesinSAFTA.¹

Tariff reductions under SAFTA will not be enough to reduce informal trade in South Asia. Official trade data are widely believed to grossly understate the “true” size of intraSouth Asian trade, given the substantial informaltradeflows.Indeedinformalimportsofwheatandwheatflourfrom PakistanensuredamoreorlesscontinuingsupplyinAfghanistanduring2007 08 despitetheofficialexportbanimposedbyPakistan.Aninitialattemptto model informal trade suggests that SAFTA has only a limited impact on informaltradeflowsacrossallcountries.Tariffreductions,intheabsenceof other institutional reforms and enforcement, would most likely have little impact on illegal crossborder trade, especially between Pakistan and Afghanistan.

LESSONSLEARNEDANDTHEWAYFORWARD

Thefoodcrisisisbynomeansover.Domesticpricesofbothwheatandrice remainhighthroughoutSouthAsia.Thereisgrowingagreementthatatwo track approach is required, combining increased investments in safety nets withmeasurestostimulatebroadbasedagriculturalproductivitygrowth,with majoremphasisonthemajorfoodstaples.

The degree of price transmission is an important determinant of consumerwelfare.Forobviouspoliticalandsocialreasons,mostSouthAsian governmentsarelikelytocontinuetoseektoprotectconsumersagainstprice variability. This requires careful management of price transmission through trade, pricing, and stocking policies, supplemented by social protection programs.

Policies and programs for managing price transmission need to be appropriately designed. Trade policies should encourage the operation of theprivatesectorandnotrestrictexports.Pricingpoliciesmayincludelimited subsidies targeted to the poor, but general control measures should be avoided.Publicgrainreservesshouldbelimitedinsize,andaninternational coordinated global food reserve—in which countries’ own reserves would becomepartofalargerglobalreserve—deservesconsideration.

Protectingconsumers’welfareandmaximizingfoodsecurityinasustainable andfiscallyaffordablewayisonlypossibleifsimultaneousattentionisgivento appropriatesupplyresponsemeasuresthatprotectproducers’welfareaswell.

1

Thesesimulationresultsdonotmeanthatexportrestrictionsimposedbyindividual countries do not matter. By restricting supplies, export restrictions can seriously augmentfoodpriceinflationinimportingcountriesthatimportalargeportionoftheir foodsuppliesfromthecountriesthatimposedtheban.

Higher food prices are not unequivocally bad and may provide new opportunities. Besides the potential benefits to net selling households and theireffectsonsupply,higherfoodpricescouldgenerateanumberofother benefits. In South Asia, they provide an opportunity to policymakers to re examine the complex system of inputoutput pricing interventions; reduce spending on input subsidies and instead refocus public spending on investments to raise farm productivity (irrigation, rural roads, electricity) as wellasonimprovedsocialprotectionmeasures.Higherfoodpricesmayalso stimulate innovative developments in food aid, in particular a shift from traditionalfoodaidtofoodassistancethroughlocalfoodpurchasescombined withcashtransfersandvouchers.Sustainedhigherfoodpricescouldalsohelp the implementation of responsible international trade policies that benefit lowincome countries, and help to reform developed countries’ agricultural supportprogramsinawaythatmayremovetheremainingbarrierstoprogress ontheWTODohatradenegotiations.

Thelongtermchallengetoproduceenoughfoodhasnotdisappeared. The underlying problems remain of low stockpiles, rising demand mainly fuelled by continuing population growth in developing countries, and flattening yield growth. These problems are particularly relevant for South Asia,giventheregion’shighpopulationgrowth.

RaisingproductivityisnecessarytoensureSouthAsia’sfoodsecurity.

Giventhatmostproductivelandisalreadyundercultivation,futureincreases in agricultural production in the region will need to be based on yield increases.Becauseworldpricesofenergyandfertilizerareexpectedtoremain substantiallyhigherthanbefore,yieldincreasesaretheonlysustainableway to reconcile higher input costs and farmers’ incentives with low and stable consumerpricesofwheatandrice.

Yieldincreasesseementirelyfeasiblegiventhesubstantialyieldgapsin

SouthAsianagriculture.Despiteafewimportantexceptions,theimpactof higherpricesoncropyieldshasbeenlimitedsofar.Toraiseyieldsrequiresa combination of technical interventions and socioeconomic policies and measures. But besides technology transfer, policymakers should ensure that the global economic crisis does not jeopardize public investment in agricultural research and rural infrastructure. Governments must also allow price incentives to reach farmers, The should ensure that adequate mechanismsareinplaceforsupplyingqualityinputsataccessiblepricesand thatfarmershaveappropriatemarketingopportunities.Inthiscontextpublic spendingonirrigation,ruralroads,andelectricityiscrucial.