Professional Documents
Culture Documents
Milk, Fodder, and the Green Revolution: The Case of Mixed Farming in the Pakistan Punjab
Author(s): Takashi Kurosaki
Source: The Pakistan Development Review, Vol. 35, No. 4, Papers and Proceedings PART II
Twelfth Annual General Meeting of the Pakistan Society of Development Economists
Islamabad, December 14-16, 1996 (Winter 1996), pp. 537-548
Published by: Pakistan Institute of Development Economics, Islamabad
Stable URL: http://www.jstor.org/stable/41259981 .
Accessed: 22/06/2014 06:31
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Pakistan Institute of Development Economics, Islamabad is collaborating with JSTOR to digitize, preserve and
extend access to The Pakistan Development Review.
http://www.jstor.org
Milk,Fodder,and theGreenRevolution:
The Case ofMixed Farming
in thePakistanPunjab
Takashi Kurosaki
1. INTRODUCTION
The Punjabagriculture, bothin India and in Pakistan,has been knownas the
centreof theGreenRevolutionin SouthAsia. Due to itshighlydevelopednetwork
of publicirrigationcanals,highyieldingvarietiesof wheatand rice wereadopted
rapidlyin thelate 1960s and theearly1970s,resulting in a phenomenalincreasein
cerealproduction.The Punjabregionshave becomethegranaryof each country,
supplying surplusfood-grains tootherparts.
A recentphenomenon thatdeservesattention is an increaseof theimportance
of dairyproduction in thePunjab agriculture [Kurosaki(1995a)]. It has occurred
bothat themacroeconomic levelas an increaseof thelivestocksharein agricultural
value-addedand at themicroeconomic levelas an increaseof themilkincomeshare
inruralhouseholdeconomy.
Despiteits importance, veryfewstudieshave investigated a microeconomic
mechanismunderlying this phenomenon. To analysethe micromechanism,it is
necessaryto pay due attention to thefactthatthemajorpartof dairyactivitiesare
carriedoutin thebackyardoffarmswherecropsand dairyproduction are combined.
AssociateProfessor,
TakashiKurosakiis currently of EconomicResearch,Hitotsubashi
Institute
University,
Tokyo.
2. STUDY AREA
The modelinthispaperis calibrated formixedfarming householdsin therice-
wheatzone of thePakistanPunjab. Parameters are obtainedfromhouseholddata of
productionand consumption, collectedin theSheikhupura
originally by the
district
Punjab Economic Research Lahore.
Institute, The data set comprises97 household
observations each forthreeyearsfrom1988-89 to 1990-91,forwhichKurosaki
(1995a) providemoredetailedinformation.
Table 1 summarises informationon cropand livestockproduction and income
sourcesforthe samplehouseholdsin 1990-91. Major cereal crops are Basmati
paddy(majorcash crop) in theKharifseason and wheat(staplefood crop) in the
Rabi season. In additionto thesetwo cereal crops,mostfarmers keep livestock
a
animalsand allocate significant share of land
cultivated to foddercrops used as
-
greenfodder mostly,Jowarin Kharifand Berseemin Rabi. The sum of areas
devotedto thesefoddercrops and the dominantcereal crops (rice in Kharifand
wheatin Rabi) is 80 to 90 percentin thestudyarea. Milk is themostimportant
livestockproductsold to marketsregularly.Mosthouseholdskeepseveralcows and
she-buffaloes for milk production.Householdsfeed livestockanimalson green
fodderfromfarmland,dryfodderfromcrop byproducts such as wheatand rice
straw,andconcentrate feedssuchas cottonseedcake. Amongthese,theexpenditure
on greenfodder(includingunpaidcosts imputedfromvillageprices)occupiedthe
largestshareofthetotalfeedcostat around70 percent.
'in the traditional before,the provisionof draftpowersby bullockto crop
Punjab agriculture
was moreimportant.However,thisfunction
cultivation is becomingless importanttodaydue to rapid
development oftractorrentalservicemarkets.
Table 1
Land and LivestockCharacteristics
and HouseholdIncome
ofSampleHouseholds(1990-91)
Mean Std.Dev.
Land (Acres)
TotalOperatedLand 11.198 9.363
TotalOwnedLand 9.036 8.724
AcreageunderBasmati 6.595 6.375
AcreageunderWheat 6.888 5.830
AcreageunderKharifFodder 1.916 1.188
AcreageunderRabi Fodder 1.825 1.285
LivestockSize
NumberofAdultShe-buffaloes 2.979 1.014
NumberofAdultBullocks 0.856 1.173
AU #ofTotalMilchAnimals 5.693 2. 109
AU #ofTotalDraftAnimals 1.389 3.506
AU #ofTotalLivestockAnimals 7.082 4.150
HouseholdIncomebySources (NominalRupees)
LivestockIncome (1) 16978 10751
CropIncome (2) 24978 20493
FarmIncome (3)=(l)+(2) 41957 24903
Off-farmIncome (4) 8122 3514
TotalHouseholdIncome= (3)+(4) 50079 23034
Source: The author'scalculation.The originalinformationwas collectedby the Punjab Economic
ResearchInstitute.
Note: The numberofobservations is 97.
#The adultanimalunits(AU) forlivestockanimalsaredefinedas follows.Draftanimals:1.0 for
0.57 foradultdonkeys,0.28
0.57 foryoungbullocks/he-buffaloes,
adultbullocks/he-buffaloes,
foryoungdonkeys,and 1.0 foradulthorses. Milchanimals:1.28 foradultshe-buffaloes,0.96
0.72 foradultcows,0.54 foryoungcows,and0.20 foradultgoats.
foryoungshe-buffaloes,
of green fodder are the most volatile among major agriculturalcommodities in the
area, suggestingthe localised natureof greenfoddermarkets. From the viewpointof
risk-aversefarmers,this structureof green fodder markets implies high risk of
transactions.Data indicatethathouseholds were well aware of the existenceof green
foddermarketsand participatedin themif necessary,but only marginally.
3. A HOUSEHOLD MODEL
A household model in this paper is based on Kurosaki (1996). Its unique
featureis thathouseholds' crop choices are affectednot only by theirwillingnessto
bear risk but also by their ordinal consumptionpreferencesfor individual goods,
whenboth income and consumptionprices are stochastic.The household model is an
attemptto link the intertemporalconsumption model of risk-coping mechanisms
(insurancein a broad sense) withtheriskmanagementproductionmodel in whichthe
distributionof household income is endogenously determined [Morduch (1990,
1995); Rosenzweig and Wolpin (1993); Rosenzweig and Binswanger(1993)].
Households' objective functionis v(K, p' ß, y, V)> where Y is the household
income definedbelow, p is a price vectorof consumptiongoods, and ß, y, and '|/ are
parameters characterising the function v. Vectors ß and y represent ordinal
consumption preferences for individual goods such as the income and price
elasticities of demand for a commodityand the share of the budget spent on the
commodity.'|/ characterises households' observed risk aversion. As discussed in
Kurosaki (1996), when households have no ex post risk coping measures contingent
on a realisation of stochastic elements,the functionv(..) is reduced to an indirect
utilityfunctionand '|/ is equivalent to households' actual risk aversion (actual risk
attitudes);when households have some ex post risk coping measures,v(..) should be
interpretedas an approximationof a households' value functionderived fromtheir
stochastic dynamic optimisation and '|/ reveals the mixture of households' risk
attitudesand the availability of these measures [Kurosaki (1996); Rosenzweig and
Binswanger(1993)].
In the model, households choose the acreage shares of the fourmajor crops in
the studyarea (Basmati paddy, wheat,and Kharifand Rabi foddercrops). They are
denoted /v/,where subscriptss (=/:, r) representtwo croppingseasons of Kharifand
Rabi, and i (=1, 2) representcereal and foddercrops. These crop activitiestogether
withmilkproductionin Kharifand Rabi, and non-farmincome YNthatis assumed to
be non-stochastic,constitutehouseholds' income flow Y as:
4. SIMULATION METHODOLOGY
The strength of structurally householdmodelis thatwelfareimpacts
estimated
of changesin exogenousparameters consistent
can be evaluatedin a theoretically
to a morecommon,reduced-form
way. This is in sharpcontrast approach,whichcan
derive only supplyresponse. In this paper, welfareeffectsare evaluated by
equivalentvariation.The welfarestatusunderregime£,is expressedas E[v(f(l¡ ,
еД Р (£/))]>wherehouseholdincomeД..) is determined by theoptimalcropchoice
of a
vector £,. Denoting lump-sum
(/,*)anda stochastic transfer
byT,supplyresponse
of theoptimalproduction choice(A/*),a changein expectedincome(А £(/"))> and
2Theestimatedversionadoptsa specificationof thelinearexpenditure system(LES) forordinal
constant
preferences, relativeriskaversion(CRRA) with respectto the LES real incomethatallows
in y accordingto households'abilityto smoothconsumption
differences
structural ex post,and quadratic
fortheproduction
functions constraints.
equivalentvariation
( xE ) ofa changefromtheinitialregimee0 to a newregime£|
aredefinedas
А/**/Г(е,)-/о(ео)
áEiY^mElftf.eO]-E[fVle0)].
£[v(/(/;,eI)fp(eI))]= £[v(/(Zo*,eo)
+ xEtp(eo))] (3)
where
Equivalentvariation
xEfora risklessenvironment,
forexample,is interpretedas the
amountof moneythatshouldbe givento a farmer undertheinitialriskyregimeto
makehimindifferent betweenthe two regimes.Therefore, it gives the monetised
valuationof thewelfarecostofrisk. Thesemeasuresof supplyresponseand welfare
changesareestimated usingnumerical whosedetailsare givenin Kurosaki
methods,
(Forthcoming).
Table 2
Resultsfor WelfareCostsofRisk
Simulation
CompleteRisk GreenFodder
Elimination PriceElimination
"AVG" Household
1. SupplyResponse(ChangeinArea)
BasmatiPaddy 29.2% 11.3%
KharifFodder -37.0% -13.3%
Wheat 47.0% 32.7%
Rabi Fodder -100.0% -66.9%
2. ChangeinExpectedIncome(EY)
% to theInitialEY 2.0% 1.6%
3. WelfareChange(EquivalentVariation)
% to theInitialEY 7.9% 4.9%
"Land-Poor" Household
3. WelfareChange(EquivalentVariation)
% to theInitialEY 10.9% 9.2%
"Livestock-Poor"Household
3. WelfareChange(EquivalentVariation)
% to theInitialEY 12% 1.9%
Notes: "AVG" is thereference householdgroupwithmediancharacteristics amongsamplehouseholds.
in greenfodderand almostalwayshavesurplusin wheat
Itis likelyto be close to self-sufficiency
and Basmatipaddy. "Land-Poor"is thehouseholdgroupwithhalfthesize of operationalland
and othercharacteristics remainingthe same. This groupis a purchaserof greenfodderon
averageand sometimesis a purchaser of wheatforfamilyconsumption."Livestock-Poor" is the
householdgroupwithhalfthesize oflivestockherdand othercharacteristics remainingthe same.
Itis a netsellerofgreenfodder, wheatandBasmatipaddy,on average.
Table 3
ofChangesinMarketStructure
Effects
Experiment (1) (2) (3) (4) (5)
"AVG" Household
1. SupplyResponse(ChangeinArea)
BasmatiPaddy 0.7% 12.3% 2.6% 3.9% 14.1%
KharifFodder -0.8% -14.6% -3.0% -4.5% -16.9%
Wheat 0.2% 8.2% 0.7% 1.1% 8.7%
Rabi Fodder -0.4% -15.5% -1.2% -2.1% -16.5%
1. SupplyResponse(Changein
OutputQuantity)
BasmatiPaddy 0.7 12.3% 2.6% 3.9% 14.1%
KharifFodder -0.8% 2.5% -3.0% -4.5% -0.2%
Wheat 0.2% 8.2% 0.7% 1.1% 8.7%
Rabi Fodder -0.4% 1.4% -1.2% -2.1% 0.2%
2. ChangeinExpectedIncome(EY)
% to theInitialEY 1.3% 6.4% 11.5% 17.2% 19.2%
3. WelfareChange(Equivalent
Variation)% to theInitialEY 2.7% 7.4% 4.4% 13.5% 14.7%
4. InducedChangesinExpectedGreen
FodderPrices
KharifFodder 3.5% -21.1% 24.6% 18.5% 1.9%
Rabi Fodder 2.2% -14.7% 12.3% 10.6% -2.2%
"Land-Poor" Household
3. WelfareChange(Equivalent
Variation)% to theInitialEY 3.8% 10.6% -4.3% 6.5% 9.5%
"Livestock-Poor"Household
3. WelfareChange(Equivalent
Variation)% to theInitialEY 1.5% 3.8% 12.5% 18.7% 18.4%
Notes: In Column(1), priceelasticitiesof greenfodderdemandin Kharifand Rabi are doubled. In
Column(2), expectedyieldsof Kharifand Rabi foddercropsare increasedby 20 percent. In
Column(3), expectedpricesof Basmatiand wheatare increasedby 20 percent.In Column(4),
thesepricesare increasedsimilarly,simultaneouslywithdoublingprice elasticitiesof green
fodderdemand in Kharifand Rabi. In Column (5), these prices are increasedsimilarly,
withincreasesinexpectedyieldsofKharifand Rabi foddercropsby20 percent.
simultaneously
REFERENCES