international price transmission in cge models: how to ... · international price transmission in...
TRANSCRIPT
1
InternationalPriceTransmissioninCGEModels:HowtoReconcileEconometricEvidenceandEndogenousModelResponse?
KhalidSiddig
AgriculturalandFoodPolicyGroupUniversityofHohenheim
Stuttgart70593Germany
PhoneNumber:(+49)711459‐22643Email:[email protected]
and
HaraldGrethe
AgriculturalandFoodPolicyGroupUniversityofHohenheim
Hohenheim70593Germany
PhoneNumber:(+49)711459‐22631Email:grethe@uni‐hohenheim.de
Contributed Paper at the 15th Annual Conference on Global Economic Analysis
"New Challenges for Global Trade and Sustainable Development"
June 27 to 29 2012, Geneva
Copyright 2011 by the authors. All rights reserved. Readers may make verbatim copies of thisdocumentfornon‐commercialpurposesbyanymeans,providedthatthiscopyrightnoticeappearsonall such copies. The authors acknowledge the generous support of the DeutscheForschungsgemeinschaft(DFG)forthisresearchprojecton“TheEconomicIntegrationofAgricultureinIsraelandPalestine”.
2
InternationalPriceTransmissioninCGEModels:HowtoReconcileEconometricEvidenceandEndogenousModelResponse?
Abstract
The importanceofempiricallyanalyzing thetransmissiontoand impactsof international
pricesondomesticmarketsisgrowing,particularlysincethe2006–2008foodpricehike.
However, the field is dominated by econometric time‐series analysis (PTA) and rather
disconnected from analyses based on simulation models such as computable general
equilibrium(CGE)models.Themissingreconciliationamongthesetoolscouldbedueto:
PTAoftenbeingconcernedwithhighfrequencydataandshorttermadjustmentprocesses,
whichdoesnotreconcilewellwiththeannualdataofCGEanalyses;typicallylittleoverlap
between research teams ineconometric time series analysis and those inCGEmodeling;
and the endogeneity of price transmission in CGEmodels. Due to this endogeneity, the
calibration of CGE models to empirically observed price transmission is not
straightforward,asaninfinitecombinationofmodelparametersandspecificationsallows
forreachingacertainlevelofpricetransmission.Thispaperaimstoaddressthequestion
of how a certain degree of price transmission from the international to the domestic
market,whichmaybedeterminedempiricallysuchasbasedonavectorerrorcorrection
model, can be met in a single country CGE model. We examine and validate seven
hypothetical determinants including structural characteristics of the model, the
parameterization of behavioral functions, and properties of the sectors concerned. The
findingsofthispapersupportcontrollingthepass‐throughofpricesfromtheinternational
tothedomesticmarketinCGEmodels.
Keywords: Price transmission, sensitivity analysis, CGE models, parameterization,
internationaltrade.
JELClassification:C13,C68,D58,F17,F31,J60.
3
1 Introduction
The literature on the empirical analysis of the transmission of international prices to
domestic markets is dominated by specific analytical methods such as cointegration
analysisandtheestimationoferrorcorrectionmodels.However,connectiontotheworld
ofComputableGeneralEquilibrium(CGE)modeling,despitetheirpowerfulnessinshowing
the economy‐wide implications of price changes, is limited. At most, prior studies have
includedexogenousassumptionsonpricetransmissionwhicharepluggedintoCGEmodels
by fixingdomesticprice levels(see, forexample,MundlakandLarson(1992),Baffesand
Gardner(2003),Delgadoetal.(2004),andHabermeieretal.(2009)).1
Adam (2011) illustrates the realmacroeconomic and distributional effects of alternative
fiscalandtradepolicyresponsestofoodpriceshocksandhowthesearedeterminedbythe
structural characteristics of low‐income economies. He also investigates impacts of food
price shocks on the aggregate price level and the implications of increased food price
volatility on the design and conduct of monetary policy in low‐income economies.
However,despitetheconsiderationofmanystructuralcomponentsthatmightimpactthe
transmission process. For example, the elasticity of substitution between imports and
domesticgoodsandtheelasticityoftransformationbetweenproductionforthedomestic
and the export market are both set to 0.5 without considering the impact of these
parametersonpricetransmission.Indoingso,thecommonapproachofselectingelasticity
values based on educated guesses and knowledge about particular countries without
takingpricetransmissionimplicationsintoaccountisfollowed.
Themissing reconciliation among the research fields of econometric price transmission
analysis(PTA)andCGEanalysiscouldbeduetoseveralreasons,suchas:(1)EmpiricalPTA
being often more concerned with high frequency data and short term adjustment
processes,whichdoesnotreconcilewellwiththetypicalsolutionperiod(annualaverages)
and simulation horizon (medium to long term) of CGE analyses; (2) Different research
teamsforeconometrictimeseriesanalysisandCGEmodelingwithtypicallylittleoverlap;
and(3)PricetransmissionisendogenoustoCGEmodelsandisdeterminedbyawiderange
1SeeAdam(2011)foracomprehensivesurveyandcomparisons.
4
ofmodelparametersandspecifications,suchastradeshares,shareofthesectorconcerned
intheeconomy,Armingtonelasticities,elasticitiesofsubstitutionamongvalueaddedand
intermediateinputs,andelasticitiesofsubstitutionswithinintermediateinputcategories,
factormarket closures in general, and their complex interactions. As a consequence, the
calibration of CGE models to empirically observed price transmission is not
straightforward;aninfinitecombinationofmodelparametersandspecificationsallowsfor
reachingagivenlevelofpricetransmission.
Thispaperaddressesthequestionofhowacertaindegreeofpricetransmissionfromthe
internationaltothedomesticmarket,whichmaybedeterminedempirically(e.g.,basedon
a vector error correction model), can be met in a single country CGE model. To this
purpose,weemployas “laboratory”a single countryCGEmodeldevelopedbyMcDonald
(2009)andadjustedtoasocialaccountingmatrix for Israel(Siddigetal.,2011).Wefirst
formulateaprioriassumptionsonwhichmodelspecificationsandparametersinaCGEare
themain determinants of the resulting degree of price transmission. Subsequently, in a
systematic sensitivity analysis we analyze the impact of these determinants and their
interactionaswell as showhowexogenouslygivendegreesofprice transmissioncanbe
met endogenously in a CGEmodel based on the calibration of various parameters. As a
simple and significant shock,we apply a doubling of international cereal andother crop
prices, which is not far from what was observed during the years 2007 and 2008, and
investigatehowthisshockimpactsdomesticpricesintheIsraelieconomy,whichhashigh
importsharesoftheseproducts.
Thepaperisorganizedasfollows:inSection2wedescribethemodelincludingitsclosure
rulesandthepricesystem; inSection3wehighlightthedatabaseincludingtheSAMand
behavioralelasticitiesandalsodescribetheselectedfoodsectorsinIsrael;inSection4we
formulate a priori expectations about the main determinants of price transmission; we
presentresultsofasystematicsensitivityanalysisinSection5;andSection6containsthe
conclusion.
5
2 TheModel2.1 OverviewThesinglecountryComputableGeneralEquilibrium(CGE)modelSTAGE(McDonald,2009)
is used as a basis of our experiments. STAGE is a Social AccountingMatrix (SAM) based
model with amix of non‐linear and linear relationships that depict the behavior of the
economy’sagents.Itisastaticmodelwithhouseholdsmaximizingtheirutilityaccordingto
preferences represented by Stone‐Geary utility functions. Households consume
commoditiesavailableinthedomesticmarketfrombothdomesticproductionandimports.
The substitution between domestic and imported goods is governed by the Constant
ElasticityofSubstitution(CES)specification(Armington,1969).Israelisasmallcountryin
theworldmarketofwheat,othercerealswhichcompriseofall cerealsother thanwheat
suchasrice,corn,barley,rye,andoats,andothercrops(thecropswhicharethefocusof
thisstudy);hence,worldmarketpricesforimportsandexportsarefixedinthemodel.
Domesticproductionismodeledasatwo‐stageproductionprocesswithaCESfunctionon
top, where intermediate inputs and production factors are combined to generate the
outputofeachactivity.Atthesecondstage,intermediateinputsarecombinedaccordingto
Leontief technology, while production factors (land, labor, and capital) are combined
accordingtoCEStechnology,withtheoptimalratioofproductionfactorsbeingdetermined
byrelativeprices.Thedomesticproductionofcommoditiesissoldinthedomesticorthe
exportmarket based on a Constant Elasticity of Transformation (CET) specification and
dependent on relative prices in these markets. The model is implemented in General
AlgebraicModelingSystem(GAMS)softwareandadaptedtoanIsraeliSAM(Siddigetal.,
2011).
TheIsraeli2004SAMisadetailedSAMthatcomprisesoneaccountforcapital,oneforland,
and36forlabor.Forthepurposeofthispaper,neitherthedifferentlaboraccountsnorthe
accounts of taxeson labor are at the focus; hence, they are discussed as one aggregated
laboraccount.
2.2 PricesystemFigure(1)showstheinterrelationshipsbetweenthepricesforcommoditiesandactivities
as depicted by the model. The supply prices of the composite commodities (PQSc) are
defined as the weighted averages of the domestically produced commodities that are
6
consumed domestically (PDDc) and the domestic prices of imported commodities (PMc).
The domestic prices of imported commodities are defined as the product of the world
prices of commodities (PWMc) multiplied by the exchange rate (ER) plus ad valorem
importduties(TMc).Weightsofthedomesticandtheimportedpricesinthedetermination
of theaveragepricesaredeterminedbyaCES function.Theaveragepricesexcludesales
taxes,andhencemustbeupliftedby(advalorem)salestaxes(TSc)andexcisetaxes(TEXc)
toreflectthecompositeconsumerprice(PQDc).
Figure(1):PriceRelationshipsfortheSTAGEModel
Source:ModifiedandextendedfromMcDonald(2009).
Theproducerpricesofcommodities(PXCc)aresimilarlydefinedastheweightedaverages
ofthepricesreceivedfordomesticallyproducedcommoditiessoldondomestic(PDDc)and
export (PEc)markets.Weights adjust endogenously based on a CET function. The prices
receivedontheexportmarketaredefinedastheproductofworldpriceofexports(PWEc)
and theexchange rate (ER) lessanyexportdutiesdue,whicharedefinedbyadvalorem
exportdutyrates(TEc).
PWEc
PWMc
PEc
PMc
PXCc
PQSc
PXa1...
PXan
PQDc
PINTa
PVAa
PDDc
Exporttax(TEc)
PWMc*ER*(1–TM)
PWEc*ER*(1‐TEc)
Productiontaxes(TXa)
Importtarifff(TM)
SIGMA
OMEGA
SIGMAac SIGMAx
SIGMAva
w‐land
w‐labor
w‐cap
Sales,excise,andother
domestictaxes
PQSc*(1+TSc)
7
Thisstudyfocussesonasampleofselectedpricestoinvestigatethedeterminantsofprice
transmission from the rest of theworld to thedomesticmarket. The selectedprices are
producer prices of commodities (PXCc), supply prices of domestically produced
commodities on the domestic market (PDDc), supply prices of composite commodities
(PQSc),valueaddedpricesofproductionactivities(PVAa),andintermediateinputpricesof
production activities(PINTa). Elasticities are also shown in the figure as the trade CES
(SIGMA), production CES at the top level (SIGMAx), production CES at the second level
(SIGMAva),andtradeCET(OMEGA).
2.3 ClosurerulesForouranalysisweapplythefollowingstandardclosure:themodel isinvestmentdriven
with the share of investment in domestic final demand being fixed. The government
account,alltaxrates,governmentconsumptionexpendituresandtransfersareassumedas
fixed,whilegovernmentsavingsarea freevariable.For the factormarket,allproduction
factorsareassumedfullyemployedandmobileacrosssectorsinthedefaultclosure,which
reflects a long term adjustment period.We also employ amodified factor closurewhich
reflects a more medium term adjustment period and considers the amount of capital,
skilledlabor,andlandtobefixedandactivityspecific,whileonlyunskilledandsemi‐skilled
laborare fullymobileand fullyemployed.Thecurrentaccountbalance isassumed tobe
fixed,whiletheexchangerateisthefreevariabletheindefaultclosure.Importandexport
pricesarealwaysfixed.
3 TheDatabase3.1 SocialAccountingMatrixTheIsraeliSAMthatrepresentstheIsraelieconomyintheyear2004isthemaindatabase
usedinthisstudy.Itincorporates43sectors,36laboraccounts,10householdgroups,and
18taxcategoriesotherthantaxesonproductionfactors(Siddigetal.,2011).TheSAMwas
developed based on data obtained from different official sources in Israel including the
IsraeliCentralBureauofStatistics(ICBS),theCentralBankofIsrael(BOI),andtheIsraeli
TaxAuthority(ITA). Inaddition,non‐Israelisourceswereusedtofill‐ingaps indomestic
reportssuchastheWorldTradeOrganization(WTO), theOrganizationforEconomicCo‐
operationandDevelopment(OECD),andtheWorldBank.
8
For this study, details pertaining to production factors, households, or taxes are not
relevant; therefore, they will be treated as aggregates throughout the discussion. What
matters from theviewpoint and for theobjectivesof this study ishow the selected food
sectors are represented in the SAM and what weight they have in the entire Israeli
economy.
3.2 SelectedfoodsectorsandpricetrendsDespitetheirimportanceinmeetingdomesticfooddemandinIsrael,wheat,othercereals,
and other crops do not represent big shares in aggregated economic variables such as
exports, imports, andproduction. Table (1), shows thepercentage share of each in total
Israeli imports, exports, production, and factor use. The three commodities together
representonly1.7%of total Israeli imports, 0.8%ofexports, andonly0.4%ofdomestic
production; however, their share in the total cultivated land is 33.1%with other crops
aloneoccupying23.4%.Thisimpliesthatthesethreecropswillbesensitivetolandmarket
closure, while changes in the three crops are expected to have a minor influence on
economicvariablesatthemacro‐level.
Table1:Sharesofthethreecommoditiesinselectedvariablesoftheeconomy(%)
Import Export Production Landuse Labouruse CapitaluseWheat 0.37 0.00 0.03 6.97 0.03 0.01Othercereals 0.59 0.01 0.03 2.87 0.01 0.07Othercrops 0.72 0.75 0.31 23.35 0.25 0.34
Total 1.68 0.76 0.36 33.18 0.29 0.42
Sources:Siddig,etal.(2011),owncalculations.
Thesizeof thesesectorsmeasuredby theircontributions to thecountry’s totaldomestic
productionshowsthatothercropshavethelargestshareofthethreesectorswitha0.31%
contribution, followed by wheat with 0.029%, and other cereals with 0.026%. Similar
ranking isalsoachievedwhenconsidering theshareof thesesectors in theiruseof total
agriculturallandandlabor.Accordingtotheircontributionstothecountry’stotalexports
andimports,thesectorothercropshasthelargestsharefollowedbyothercerealsandthen
wheat.Thereisasimilarrankingforthesharesofthesesectorsintotalcapitaluse.
If we c
smaller
totalsu
where7
the larg
supply
respons
cereals.
Nodom
experim
commo
On the
cereals
totalde
as inter
differen
0%
20%
40%
60%
80%
100%
consider to
rcontributi
upply is low
71%ofitsd
gestof the
is importe
sive to inte
.
mesticpolic
ment. More
oditiesisno
Figure
domesticd
aremostly
emand,resp
rmediate,
ntdemand
%
%
%
%
%
%
tal domest
ionthanth
west foroth
domesticsu
threeselec
ed (Figure
ernational
cyinterven
eover, the
otexpected
e(2):Comp
Sour
demandsid
ydemande
pectively.T
32% for e
structures
71.2
23.9
5.0
Wheat
tic supply,
heothercro
hercereals
upplyisim
ctedsector
2). Accordi
market pr
ntionrelate
share of t
toinfluenc
onentsofd
rces:Siddig
deof the th
edas interm
Thesituatio
exports, 20
arealsoex
Othe
9
the wheat
opssector.
swhereab
ported(Fig
rs in terms
ingly, the o
ices from t
dtotaxes,
taxes and
cetheresul
domesticsu
g,etal.(201
hreesector
mediates,w
onisdiffer
0% by hou
xpectedtor
73.6
13.8
12.6
ercereals
t and other
Theshare
out74%is
gure2).Int
ofproduct
other crops
the import
subsidies,
margins i
ltssignifica
pplyofthe
11),ownca
rs, shown i
whichaccou
rentforoth
seholds, an
resultindif
30
56
1
Other
r cereals se
ofdomest
s imported
theothercr
tion,only3
s sector is
t side comp
ortariffs i
n the sup
antly.
threecomm
lculations.
inFigure (3
unt for99%
hercropsw
nd 5% for
fferentimp
0.4
6.6
3.0
rcrops
ectors have
ticproducti
, followed
ropssector
30%of its
expected t
pared tow
isconsidere
ply of the
modities
3),wheata
%and95%
with44%d
r investmen
pactsofwor
Taxesmargi
Produ
Impor
e amuch
ioninthe
bywheat
r,whichis
domestic
to be less
wheat and
edinthis
selected
andother
%of their
demanded
nt. These
rldprices
andns
uction
rt
changes
resultin
thantho
Thecos
demand
56%, w
respect
factors.
Historic
domest
motivat
trendsi
0%
20%
40%
60%
80%
100%
sondomes
ndomestic
oseofwhea
Figure
ststructure
d bywheat
while the
tively. The
.
calandrec
ticmarketa
tedbythe
inIsrael.
99
W
sticmarket
cpricesoft
atandothe
(3):Compo
Sour
eofthethre
t, cereals, a
share of t
remaining
centinform
andthewo
importance
9.41
Wheat
ts.Forexam
thisproduc
ercereals.
onentsofdo
rces:Siddig
eesectorsd
and other
the deman
g share is
mationabou
rldmarket
eofthetra
3.
94
Other
10
mple,theh
ctbeingmo
omesticdem
g,etal.(201
doesnotdif
crops in th
nd for pri
devoted t
utpricetren
tarealsore
ansmission
.90
4.97
cereals
higherexpo
orerespons
mandofthe
11),ownca
fferverym
he total cos
mary facto
to taxes on
ndsofthe
elevanttoo
ofworldm
31.
4.9
19.
43.
Other
ortshareof
sivetochan
ethreecom
lculations.
much.Thesh
st structure
ors is 37%
n producti
threecomm
ourexperim
marketpric
.92
97
.59
.52
crops
fothercrop
ngesinexpo
mmodities
hareofinte
e is 62%, 5
%, 43%, a
ion and pr
moditiesin
ment,asthi
cesinshapi
Interm
Househ
Investm
Export
psshould
ortprices
ermediate
54%, and
and 43%,
roduction
nboththe
isstudyis
ingprices
edidate
hold
ment
Figur
Figure
percent
analysis
interna
Moreov
correlat
3.3 ElaFour se
transm
ofsubs
and int
compon
domest
2Simple
0
250
500
750
1000
1250
1500
1750
Price(NIS)
re4:Trend
4 shows th
tage chang
s, it seems
tional pric
ver, other
tedduring
asticitiesets of elast
issionfrom
titution(CE
termediate
nents, and
ticgoods.S
averageofNo
0
0
0
0
0
0
0
0
2001 2
Wheatyea
Wheatwor
sofdomest
Sources:I
hat domest
ges almost
s that the d
ce of whea
cereal and
thelastten
ticities are
mtheworld
ES)betwee
inputs, th
the consta
Section4w
o.1hardred
126
14 10
2002 2003
arlychange(%
rldprice
ticandwor
CBS(2011)
tic and int
throughou
domestic p
at represen
d wheat p
nyears.
considered
dtoIsraelt
enimports
he CES for
ant elastic
will further
winterordin
101
24
2004 20
%) C
C
11
rldpricesof
),USDA(20
ternational
ut the peri
price of wh
nted by Ka
prices in I
d crucial in
throughim
anddome
the substi
ity of tran
add to the
aryproteina
1
1
‐14
3
005 2006
Cerealyearly
Cereal
fwheatand
011),ownc
annual pr
iod. Even
heat in Isra
ansas2 whe
srael and
n the deter
portsande
sticgoods,
tution amo
nsformation
eassumptio
nd13%prote
60
14
37
1
2007 200
ychange(%)
dothercere
calculations
ices for ce
without a
ael is highl
eat in Figur
Kansas ap
rmination o
exports:th
theCESbe
ong the dif
n (CET) be
onsrelated
ein,Kansas.
‐38
17
‐22
08 2009 2
Wh
eals(2001–
s.
reals show
formal tim
y influence
re 4 (USDA
ppear to b
of the leve
econstant
etweenval
fferent valu
etween exp
d to the inf
19
35
423
2010 2011
heat(avg)
–2011)
w positive
me series
ed by the
A, 2011).
be highly
l of price
elasticity
lueadded
ue added
ports and
fluenceof
‐60
‐40
‐20
0
20
40
60
80
1
Annual%changes
12
elasticities on the transmission of the world price of imports and exports to domestic
markets,whiletheresultssectionwillassessandvalidatetheseassumptionsbasedonour
empiricalfindings.
For thedefault version of ourmodel analysis (pre‐sensitivity analysis), the valuesof the
fourelasticitiesareassumedtobe2.0forthetradeelasticities(CESandCET)and0.8for
production CESs at top and second levels. The selection of these values is based on
educatedguessesgovernedbyourknowledgeabouttheIsraelieconomyandguidedbythe
elasticityrangesrecommendedbySadoulet&deJanvry(1995).
4 APrioriAssumptionsonDriversofPriceTransmission
Apriori,wehypothesizethefollowingmodelcomponents,parameters,andspecifications
tobecrucialinthedeterminationofthedegreeofpricetransmissionbetweeninternational
anddomesticprices:
1. The trade shares (import and export): the higher the trade shares of the sector, thehigherthetransmissionofworldmarketpricechangestothedomesticmarket.Importsand exports are imperfect substitutes for domestic products. Therefore, if the initialtrade share is very small, it cannot become large (Hanslow, 2001; Kuiper and vanTongeren,2006).
2. The value share in the domestic economy: the higher the share of the sector in thedomestic economy, the higher the impact created by changes in world price on thedomestic market price. This is because the domestic price increase is dampened bymore production, which happens more easily with smaller sectors. Therefore, if thesector is large, price increases can be dampened by domestic production to a lesserextentonly.
3. Theshareindomesticfactoruse:sectorscouldbedifferentiatedaccordingtotheiruseof different factors of production with different implicit price elasticities of factorsupply.Incaseasectorreliesheavilyonaninelasticfactor,pricetransmissionislikelyto be higher. This is because with increasing domestic production, the cost ofproduction would increase more than it would if there were a higher factor supplyelasticity.
4. Thedegreeof factormobilityamongsectors: thehigherthefactormobility, the lowerthepricetransmission.Thisisbecauseincreasingdomesticsupplyincaseofincreasing
13
international prices dampens the increase in domestic prices; however, the moreimmobilethefactorsare,thelessthismechanismcanwork.
5. ArmingtonelasticitiesandCETelasticitiesbetweendomesticdemand/importdemandand domestic supply/export supply: the higher the trade elasticities, the higher theprice transmission. This is because higher elasticities allow for higher degrees ofsubstitution, resulting inmore demand for the domestically produced product in thecaseofanincreasingworldmarketprice.
6. Elasticitiesofsubstitutionamongvalueaddedandintermediateinputs:thehighertheelasticities, the more options producers have to substitute primary factors byintermediatesandviceversa.Therefore,thehighertheelasticities,thelessofanimpactworldmarket price changes will have on domestic prices. In addition, elasticities ofsubstitutionwithin input categoriesmatter because higher elasticities of substitutionenableproducerstoadjustproductionandthusdampenthepriceshockdomestically.
7. Theprevailingexchangeratepolicy:thedomesticcurrencytendstoappreciateifworldpricesofcommoditieswhicharepredominantlyexportedincrease,whilethedomesticcurrency tends to depreciate if there are rising prices for predominantly importedproducts.3 Fromaprice transmissionperspective, an increasingworldprice togetherwithaflexibleexchangeratewouldreducethepass‐throughofthatincreaseincaseofacurrencyappreciationandincreasethepass‐throughincaseofacurrencydepreciation.
Thefollowingsectiontriestoempiricallyassessandvalidatethedescribedhypotheses.
5 ResultsandSensitivityAnalysis
Asafirststep,weapplyashockofa100%increaseinworldpricesforwheat,othercereals,
and other crops to the model with standard parameterization and model closures as
described in Section 2. Thismodel is found to transmit the 100% increase in theworld
pricetothedomesticpricesofthethreecommoditiesatverydifferentmagnitudes(Figure
5). Theprices shown inFigure5 include thedomesticpriceof imports (PM), the supply
priceof the composite commodity (PQS), the supplypriceof thedomestic commodity in
the domesticmarket (PDD), the composite producer price (PXC), the value added price
(PVAa),andtheintermediateinputprice(PINTa).
3 The effects of international price shocks on the exchange rate in CGE models depend on the complexinterplayofvariousfactors(Devarajanetal.,1993)andarenotdiscussedinthisarticle.
14
Asafirstobservation,thepercentageincreaseinPMandPEis96%,ratherthanthe100%
increase in theworldmarketprice.The increase isnotby the full amountbecauseofan
appreciationofthedomesticcurrencyagainstthedollarbyabout1.8%.Whilethechange
of thepricesof importsandexports isequaloverallproducts, the transmission toother
domestic prices differs heavily among products. For example, producer prices of wheat,
other cereals and other crops increase by 25%, 32%, and 74%, respectively, due to the
100% increase in their world market prices (Figure 5). We make an effort to explain
differences among products in Section (5.1) based on sectoral shares in the domestic
economyaswellastradesharesintotalproduction/use.
Given the high level of observed correlation between the world market price and food
prices in Israel (Figure 4), we find that domestic price changes in the standard model
specificationaretoolow.Therefore,wecarryoutasensitivityanalysisinSections5.2to5.4
to investigateoptionsneeded to reachahigher level ofprice transmissionand to assess
andvalidatethehypotheticalassumptionsmadeinSection4.
5.1 SectorsharesThefirstapriorihypothesisassumesthatthehigherthetradeshare,thehighertheprice
transmission.Thereactionoftheselectedsectorstotheworldpriceincreasemeasuredat
thedomesticsupplyprice(PQS)ranksthesectorsaccordingtotheirimportsharesasother
cereals,wheat,andothercropshave73%,69%,and65%increasesinPQS(Figure5)and
importsharesare74%,71%,and30%(Figure2),respectively.Forothercrops, the65%
increase inPQSseemshigh ifoneconsiders thesubstantially lower importshare for this
product group. The impact on PQS for other crops, however, also stems from a strong
increaseintheexportprice.
Toconfirmourinterpretationregardingthecausesofpricechangesandparticularlythose
onPQS,wesimulatedasituationwheretheshocksareintroducedinseparateexperiments,
once from the export side alone and once from the import side alone. In case we only
increaseimportprices,butnotexportprices,theshareincreaseinPQSforothercropsis
indeed significantly lower: only 54%. On the other hand, wheat and other cereals are
virtuallynotexportedandtheeffectonPQSstemsalmostexclusivelyfromtheimportprice
shock.Ifweapplythepriceshockontheexportsideonly,effectsonPQSareclearlyranked
accordi
third.
Figur
Foroth
isstron
impact
forothe
price in
product
increas
pricesf
strictly
Inconc
trade s
shares
respect
termso
theresu
5
10
15
20
25
30
%changeinprices
ingtoexpo
re(5):Chan
herprices,t
ngestforoth
oftheexpo
ercropsmo
ncrease an
ts. Value a
es in produ
forinterme
followthei
clusion,we
hares. Base
of sectors
tand theo
ofitssector
ults.
96 96
0
0
0
0
0
0
0
PM
rtshares,w
ngesindomwithdefau
thepicture
hercrops,w
ortpriceon
otivatesdom
d results i
added and
uction,wh
ediates.The
irPXCcoun
canattribu
ed on this
in domest
thercrops
ralsize,bu
69
6 96
M
Wheat
withother
mesticpricesultmodelp
islesshom
whichcanb
nPXC(see
mesticprod
in a higher
aggregate
ich transla
erefore,the
nterpartsan
utethesize
single exp
tic product
sectordiff
utalsointe
73 65
PQS
t
15
cropsrank
sfollowingarameteriz
mogeneous.
beexplaine
Figure1).T
ducerstop
r increase
e intermedi
ates intohi
resultsfor
ndrankthe
eofthedom
periment, w
ion, aswh
fersheavily
ermsofits
24 2845
PDD
Otherce
kingfirst,ot
a100%inczationandc
.ForbothP
edbythehi
Thestrong
producemo
in its valu
iate input
gherwage
rPVAandP
ethreesect
mesticpric
we cannot
eat and ot
y fromthe
tradestruc
25 32
7
PXC
ereals
thercereals
creaseininclosures(%
PDDandPX
ighexport
increasein
ore.Thisda
ue added p
demand is
s for factor
PINTareno
torssimilar
ceeffectsto
conclude t
her cereals
other two
cture,both
89
12
4
PV
Oth
ssecond,a
nternationa%)
XC,thepric
shareandt
ntheprodu
mpensthe
price than
s always d
rs aswell
otsurprisin
rly.
oimportan
the relevan
s are simila
osectorsno
ofwhichi
1
24
253
VA
hercrops
ndwheat
lprices
cechange
thedirect
ucerprice
domestic
for other
driven by
ashigher
ng,asthey
ndexport
nce of the
ar in this
otonly in
mpacton
3 8
PINT
5.2 FacToanal
worldp
thoseo
level of
validate
thethre
Otherc
undert
theleve
othercr
and67%
Figure(
Wefind
price(P
worldm
theexp
whichw
strong
7
‐50
0
50
100
150
200
250
300
ctormobilyzetherel
priceshock
obtained in
f price tran
edbyourr
eecommod
cropsandc
themobilit
eloftheva
ropsandce
%,respecti
(6):Change
d that the c
PDD)foroth
marketpric
portpricefo
would be a
increase in
7469
57
57
256
‐1
Mobile
Wh
ilityevanceoff
kunderthe
amodel c
nsmission i
esults,show
ditiesunder
cerealsinp
tyandimm
alueaddedp
erealsunde
velyunder
esinprices
composite
hercropsw
ceunderth
orothercr
able to attr
nproductio
97
71
28
29
108
Immobile
heat
factormobi
eassumptio
losurewith
is lowwhe
wninFigur
rtheassum
particularsh
mobilityass
prices,whe
ertheimmo
rthemobili
withfactor
supplypri
wouldeven
eassumpti
ropscauses
ract produc
onand thus
74
54
15
18
6
1
e Mobile
Ot
16
ilityforthe
onoffullfa
h full facto
en there is
re(6),wher
mptionoffa
howhuged
umptions.
eretheper
obilityassu
tyassumpt
rmobilityv
ce (PQS) a
ndeteriora
ionoffacto
sastrongi
ction factor
ssupply fo
97
74
29
3
67
‐1
e Immo
hercereals
edegreeof
actormobil
r immobili
high facto
retheperc
actormobil
differences
Asexpecte
rcentagech
umption,re
tion.
ersusimmo
and thedom
teinrespo
rmobility.
ncreasein
rs and incr
or thedome
74
‐1
‐2
33
133
3
obile Mo
pricetrans
lityandcom
ity.Ourexp
rmobility.
centagecha
lityandim
sinthedeg
ed,thehigh
angesare2
espectively,
obilityassu
mestically s
onsetoa10
Thisisbec
thecompo
reaseprodu
esticmark
97
27
40
139
‐2
obile Im
Othercro
smission,w
mparethe
pectation is
This expe
angesinthe
mmobilityar
greeoftran
hestreactio
267%and
,compared
umptions(%
soldprodu
00%increa
causethein
ositeprodu
uction four
etwouldr
71
54
77
267
9
mmobile
ps
werunthe
resultsto
s that the
ctation is
epricesof
reshown.
nsmission
onsareat
133%for
dto139%
%)
uct supply
aseinthe
ncreasein
ucerprice,
rfold.The
esult ina
PM
PQS
PDD
PXC
PVA
PINT
17
fallofPDDby27%.PQSwouldbeaffectedbyboththisdeclineinPDDandtheincreasein
PM,withaslightlynegativenetresultof‐1%.
Someofthewheatpricesbehavecontrarytoexpectations;namelyforPVA,PXC,andPDD,
the pass‐through from the international to the domestic market is less under the
immobilityassumption than it isunder themobilityassumptionbecauseof crosseffects.
With mobile factors, the production of other crops increases heavily and replaces land
whichhadpreviouslygrownwheat:landuseforwheatdeclinesby81.4%.Thisresultsina
scarcedomesticsupplyofwheatandthushigherdomesticwheatpricesthanwouldoccur
undertheassumptionoffactorimmobility.
5.3 SubstitutionelasticitiesToanalyzetheimpactoftheproclaimeddeterminants(5)and(6),weperformasensitivity
analysis with several values for the following major related elasticities: the trade
elasticities (the CES between imports anddomestic goods and the CETbetween exports
anddomesticgoods)andtheproductionCESs(theCESbetweenaggregateprimaryfactors
andaggregateintermediateinputsandtheCESbetweenthedisaggregatedprimaryfactors
ofproduction).
Toenhancethelevelofpricetransmission,thedefaultfactormarketclosureisadjustedfor
thissensitivityanalysis.Weconsidertheamountofcapital,skilledlabor,andlandusedto
be fixedandactivity specific,whileunskilledand semi‐skilled labor are fullymobile and
fullyemployed.
Dependingonthesensitivityofthemodeltothefourelasticitiesanditsabilitytoconverge,
tendifferentvaluesareselectedforthetradeCESsandCETscoveringarangebetweenT1
(0.5) and T10 (9.5), while assuming similarity between both of them at each stage. In
addition, another ten values are selected for the production CESs assuming similarity
between both of them in each stage. The range for the production CESs is between P1
(0.05)andP10(3.0).Theexactvaluesofthetradeandproductionelasticitiesaswellasthe
defaultlevelsusedinpreviousexperimentsareshowninTable(2).
18
Table(2):Therangesofelasticityvaluesconsideredinthesensitivityanalysis
Codes ProductionCESs Codes TradeCESandCETP1 0.05 T1 0.50P2 0.38 T2 1.50P3 0.71 T3 2.50Defaultversion 0.80 2.00P4 1.03 T4 3.50P5 1.36 T5 4.50P6 1.69 T6 5.50P7 2.02 T7 6.50P8 2.35 T8 7.50P9 2.67 T9 8.50P10 3.00 T10 9.50
Figure(7)showstheresultsobtainedfromthedescribedcombinationofelasticities.Trade
CES and CET are presented at the horizontal axis, while the vertical axis shows the
percentagechangeinprices.ProductionCESsareidentifiedbydifferentcolors.
Theresultsofthesensitivityanalysisconfirmourfifthassumptionwithrespecttothetrade
elasticitiesasshowninFigure(7):allthelinesslopeupasTvaluesmovefromT1toT10.
Theexceptionsarethecompositesupplypricesofwheatandothercereals,whereprices
increasewith increasing tradeelasticitiesonly if theproductionelasticity isunityor less
(levels P1 to P4). With higher production elasticities (levels P5 to P10), higher trade
elasticities result in less price transmission. This can be explained by considering two
differenteffectswithoppositedirectionscausedbyhigherimportCESelasticitiesincaseof
increasingimportprices:
Effect1:Ahigherelasticityresultsinmoresubstitutionofthedomesticcommodityfortheimportedcommodity,whichraisesPDDandthusPQS.Effect2:AhigherelasticityresultsinahigherweightofPDDandalowerweightofPMintheformationofPQS,whichdepressesPQS.
Thelowertheproductionelasticities,thestrongereffect1(increasingPQS),andthelesser
effect2(depressingPQS).Forwheatandothercereals,effect1dominatesthetotaleffect
on PQS up to a production elasticity of up to unity, whereas effect 2 dominates with
productionelasticities aboveunity. In contrast, forother crops theeffectofhigher trade
elasticities on the price transmission from trade prices to PQS is always positive. This
19
results from the fact that PQS is also impacted by the export price due to a significant
exportshareinthisproductgroup.
The sixth a priori assumption – the lower the production elasticities are, the higher the
pass‐through is – is also validated according to our results. According to Figure (7), the
percentagechangeinthepriceseriesincreasesasproductionelasticitiesmovetowardT1
whichhasthesmallestTvalueof0.05.Thiscanbeobservedfromthewideverticalspaces
between P values, which become even wider as the elasticity is closer to P1 (see, for
instance,PXCforwheatandcereals).
Thisbeingsaid,thestrengthofpass‐throughalsodiffersamongthedifferentcommodities.
Figure (7) shows wheat prices reacting stronger to the production elasticities, which is
reflected in thewider spaces between the different lines compared to cereals and other
crops. On the contrary, other crops show the lowest reaction among the three selected
commoditiesasthelinesareclosertooneotherforallofothercrops’priceseries.Thisis
explainedbythesizeoftheothercrops’sector,ingeneral,andtheamountofresources(VA
and intermediates)needed tobringclear increases in itsdomesticproduction.Theother
cropssectoristhethirdbiggestagriculturalsector inIsraelafterthefruitandvegetables
sectorandother theanimal farmingsector,while thewheatandothercerealsectorsare
smaller.
Despitethedifferencesamongthethreecommoditiesintheirsize,coststructure,demand
components, and factor demand, their overall reaction to the elasticity combinations is
foundtobeconsistentwiththeelasticity‐relatedaprioriassumptions.Thejustificationfor
increasesinthepass‐throughastradeelasticities(CESandCET)increaseisthatdomestic
producerswould realize higher demand for their products due to domestic supply cuts
causedbyhigherworldpricesasimportsdeclineandexportsexpand.Ifdomesticgoodsin
this case are not good substitutes for imported ones, the reduction in imports and the
increaseinexportswouldhardlyhappen.
ThehighertheproductionCESis, thestrongerthereallocationamongproductionfactors
andintermediatesarewhichallowsforastrongerresponsetohigherworldprices.Thus,
thehighertheproductionCESis,thegreatertheabilityofproducerstoreactanddampen
thedomesticpriceincreaseisbyadditionalsupplyandviceversa.
ProdPercentagechangeinpricesfrom
thebase
Wheat
Othercereals
Othercrops
Figu
ducerprice(PXC)
ure7:Percenta
) Value
ageincreasein
eaddedprice(PV
TradeE
20
pricesaccordin
VA) Composi
lasticities(CES&
ngtoelasticityc
itesupplyprice(
&CET)
combinations
(PQS) Domestticsupplyprice((PDD) Productionelasticities(top&secondlevels)
21
Figure(8)helpsdetermine theobtainablerangesofprice transmissionusingdifferent
combinations of trade and production elasticities. For simplicity, only prices of other
cropsaredepicted,wheretheselectionofothercropsisrelatedtothesizeofthesector
in the economy: the other crops sector represents the largest food crop sector, with
about70%of its supplyproduceddomestically.Moreover, it couldbeconsideredasa
typical sector with substantial amounts imported, produced domestically, exported,
consumedbyhouseholds, invested,andusedas intermediate inputs.Thefigureshows
thedifferentrangesofpriceschanges indifferentcolorswhilecombiningthedifferent
trade and production elasticity values. The highest changes are always shown in red,
followedbypurple,lightgreen,orange,whilethelowestpricerangesarealwaysinblue.
Looking at the domestic prices of imports and exports, which represent the bridge
between the domestic market and the rest of the world, we find only two possible
ranges of changes in prices. The highest obtainable price change for both is 98% by
combining P1 and T10, while the lowest obtainable price change is 65% at the
combination of P10 and T10. The differences between changes in PWM and PM and
changes in PWE and PE are driven by the exchange rate, which appreciatesmore in
thesescenarioswhentradeelasticitiesarehigher.Thiswayofpresentingpricechanges
allowsfordeterminingvariouscombinationsoftradeandproductionelasticities,which
resultinthedesiredlevelofpricetransmission.
If, for example, the time series analysis confirms that a 100% increase in the world
market price of other cropswould transmit fully to the producer price, then the CGE
modelshouldbeabletogenerateanincreaseintheproducerpricecloseto100%.The
requiredelasticitylevelscanbefoundbasedonthegraphatthetoprightpartofFigure
(8),whichshows that tradeelasticitiesatT10andproductionelasticitiesatP1would
resultinthedesiredresult.However,theselectedvaluesoftradeelasticitiesarehigher
than the recommended range of elasticities that CGE modelers tends to rely upon
(Sadoulet and de Janvry, 1995).4 Thus, these results indicate the importance of
considering the specifics of individual countries and commodities in judgingplausible
4 Experience has shown that the empirical results obtained from simulations with CGEs are quiteinsensitive to specific values of elasticities, while, instead, they crucially depend upon their order ofmagnitude.Thepossiblerangeofsubstitutabilityisrelativelywellrepresentedbyfourvalues:0.3forverylow substitutability, 0.8 formedium‐low, 1.2 formedium‐high, and 3.0 for very high (Sadoulet and deJanvry,1995).
elasticit
typicall
Fig
Ifweco
rangeo
100%.T
becom
thepro
tyvalues.F
lybeonlyo
gure8:Pric
onsiderthe
ofprices is
Thelowest
mbinedwith
oducer pric
Furthermo
oneofmany
cerangesfo
eproducer
sreflected
ttradeelas
hP1(thelo
ce.On the
re,calibrat
yobjective
orothercro
rpriceofo
by thered
sticitythat
owestprod
other hand
22
tingacerta
esinthepar
opswithdi
othercrops
dcolorand
couldbec
ductionCES
d, thehigh
aindegree
rameteriza
fferentcom
sasanexa
dcomprises
onsidered
Ss)onlyto
hest produc
ofpricetr
ationofaCG
mbinations
mple,theh
schangesb
hereisT1;
generatea
ction elasti
ransmission
GE.
ofelasticit
highestobt
between8
;however,
achangeof
icity that c
nwould
ties
tainable
0%and
itcould
f84%in
couldbe
23
considered to obtain changeswithin the same range is P5, which could be combined
with T10 (the highest trade elasticities) only to generate an 80% increase in the
producerpriceofothercrops.
The second obtainable range of producer prices (60% to 80%), which is colored in
purple, couldbeachievedbycombining theproductionCESsbetweenP2andP8with
trade elasticities at different levels. As a general rule, the higher the value of the
production CESs, the higher the trade elasticitiesmust be to remainwithin the same
range of price pass through. The lowest obtainable level of pass‐through within this
range(60.8%)canbeachievedbycombingP8andT8,whilethehighest levelofpass‐
through(79.7%)canbeachievedbycombingP5andT9.
Asimilarapproachcouldbefollowedtoidentifydifferentrangesfordifferentpricesand
commoditiesinlightofthelevelsofcorrelationsbetweentheworldmarketpricesand
thedomesticpricesthroughoutthepricetransmissiontreeshowninFigure(1).Forthe
valueaddedpriceofFigure (8) (righthandmiddlepanel), forexample, thereare four
possible ranges of price pass‐through, there are three for the composite supply price
(right hand lower panel)and three for the intermediate input price (left handmiddle
panel).
These results confirm and validate our fifth and sixth hypotheses,which assume that
trade and production elasticities have significant influences on the degree of price
transmission from theworld to the domesticmarketwithin the CGE framework. It is
alsoshownthatthetransmissionishigherifweincreasethevalueoftradeelasticities,
while it is lower if we increase the value of production elasticities. Therefore, the
elasticitycombinationthatallowstheCGEmodeltogeneratethehighest levelofpass‐
through is always the one that considers high trade elasticities together with low
productionelasticities.
5.4 ExchangerateregimeToexaminetheinfluenceoftheprevailingexchangerate(EXR)regimeofthepricepass‐
through,anupdatedmodelclosureisappliedwithafixedEXR.Asdiscussedabove,the
domesticcurrencyappreciatesunderour100%worldmarketpriceshock.Thisimplies
that domestic import and export price changes would be smaller than worldmarket
pricechanges.A fixedexchangerateregime, incontrast,wouldassure thatchanges in
thedomesticpricesof importsandexportswould remain similar to changes inworld
prices as the EXRwould remain fixed at a value of 1. Figure (9) compares the pass‐
through
commo
Figur
Figure
ahighe
appreci
thepas
6 Co
Despite
prices t
empiric
cointeg
powerf
compar
regardi
actorso
Itthere
harmon
attracti
empiric
0
20
40
60
80
100
120
Changeinprices(%
)hofa50%
odities.
re9:Thepa
(9)confirm
rpass‐thro
iationand
s‐throughw
nclusion
e the grow
to domesti
calmethod
grationana
ful method
red to eco
ing implica
oftheecon
eforeseem
nizes their
iveapproa
callydeterm
49
37
1314
46
Flexible
W
increasein
ass‐through
msourseve
oughofwo
anincrease
wouldbeh
ns
wing numb
icmarkets
dsapplieda
alysisandt
ds to desc
onomy‐wid
ations a ce
omy.
sadequate
linkages in
ch is to ca
minedbyt
50
38
14
15
6 4
0
e Fixed
Wheat
nworldpr
hofa50%iex
enthaprior
rldpricein
einworld
higherinth
ber of stud
, particula
areconcent
theestimat
cribe and
e models
ertain leve
etouseav
naway th
librateCGE
imeseries
4939
15
17
47
1
d Flexib
Ot
24
ricesofexp
increaseinxchangerat
riassumpt
ncreasesto
marketpri
hecaseofa
dies analyz
rly after th
tratedone
tionofvec
forecast p
such as C
l of price
varietyofto
hat capture
Emodels t
analyses,w
50
40
16
7 1
60
2
ble Fix
thercereals
portsandim
nworldpricteregimes
tionthataf
domesticm
ices.Howe
fixedEXR
zing the pa
he 2006‐20
conometric
torerrorc
price relat
GE models
transmissi
oolsthatco
esandrefle
tomeet the
whichisou
49
33
22
18
62
2
ed Flex
mportsfor
cesunderfl
fixedexcha
marketsin
ver,ifworl
compared
ass‐through
008 food p
ctimeserie
correctionm
tionships t
s, they are
ion might
onnectsbo
ects thestr
e levelofp
urmotivati
50
34
2 2
36
108
4
xible F
Othercrop
thethrees
lexibleand
angeratere
caseofac
ldpricesd
toaflexibl
h of intern
price incre
esanalysis
models.Th
through ti
e less info
have on d
othtechniq
rengthofe
pricepass‐t
iontocond
4
23
37
110
5
Fixed
ps
selected
dfixed
esultsin
urrency
ecrease,
eEXR.
national
ase, the
suchas
heseare
me, yet
rmative
different
quesand
each.An
through
ductthis
PM
PQS
PDD
PXC
PVA
PINT
25
study.Wehypothesizesevendifferentdeterminants,basedongeneraleconomicsense
andtradetheory,ofthepass‐throughofinternationalpricestodomesticmarkets.They
aregrouped in fourmajorcategories: (1)structuralcharacteristics, suchas thesizeof
the concerned commodity in the domestic economy in terms of supply, demand,
production,andtrade,aswellasitsshareintheuseofdomesticfactorsofproduction;
(2)modelclosureandbasicassumptionsrelatedtothemobilityofproductionfactors;
(3) model parameterization, including assumptions related to the substitution
possibilities among domestic and traded goods, as well as producers’ options to
substitute primary factors by intermediate inputs or to substitute different primary
factorsforoneother;and(4)theprevailingexchangerateregime.
We use a single country CGE model developed by McDonald (2009) together with a
detailedIsraeliSAMfor2004(Siddigetal.,2011)asthebasisforourexperiments.Our
findingsconfirmthatthepricepass‐throughfrominternationaltothedomesticmarkets
in CGEs can be controlled for by configuring its closure rules, production and trade
elasticities, factor market closures, and exchange rate regime. General conclusions
includethefollowing:(1)highertradesharesofthesectorresultingreatertransmission
oftheworldmarketpricetothedomesticmarket;(2)increasingexportpricestogether
with lowCET elasticities and high production elasticitiesmay even result in negative
effectsondomesticconsumerprices; (3)greater factormobility results in lowerprice
transmission;(4)highertradeelasticitiesresultingreaterpricetransmission;(5)higher
production elasticities (top and second level) result in more substitution options for
producersandlesspricetransmission;and(6)thepass‐throughofanincreasingworld
price is higher under a fixed exchange rate regime compared to a flexible one if the
domesticcurrencyappreciatesduetotheworldmarketpriceshock.
Thedescribedapproachtodepicttheobservedtransmissionofinternationalcommodity
pricestodomesticmarketsmaybehelpfulinconsideringacombinationofPTAandCGE
modeling. The possibilities of combining the different assumptions of the model
structure,elasticities,andclosurerules tomeetacertaindegreeofprice transmission
are manifold. The modeler therefore has considerable freedom to choose a model
formulation which targets a certain degree of price transmission. This freedom is
helpful, as achieving a certain degree of price transmission will typically be just one
target among others, such as capturing the real economy accurately in the choice of
factormarketclosuresorbasingbehavioralparametersonempiricalanalysis.
26
7 References
Adam, C. (2011). On the Macroeconomic Management of Food Price Shocks in Low‐incomeCountries.JournalofAfricanEconomies,Vol.20,AERCSupplement1:63–99.
Baffes, J. and B. Gardner (2003). The Transmission of World Commodity Prices toDomestic Markets under Policy Reforms in Developing Countries. The Journal ofEconomicPolicyReform,6(3):159–80.
Delgado, C., N. Minot and M. Tiongco (2004). Evidence and Implications of Non‐tradability of Food Staples in Tanzania, 1983–1998. MTID Discussion Paper 72.Washington,DC:InternationalFoodPolicyResearchInstitute.
Devarajan, S., Lewis, J. D. and S. Robinson (1993). External Shocks, Purchasing PowerParity,andtheEquilibriumRealExchangeRate.TheWorldBankEconomicReview,7(1):45‐63.
Habermeier,K.,I.Otker‐Robe,L.Jacome,A.Giustiniani,L.Ishi,D.Vavra,T.KisinbayandF.Vazquez(2009).InflationPressuresandMonetaryPolicyOptionsinEmergingandDevelopingCountries:ACross‐regionalPerspective.IMFWorkingPaper09/1.
Hanslow,K.(2001).TheApplicationofCRESH‐BasedFunctionalFormsinCGEModels.ResearchMemorandumCatNo.:GA506,Melbourne.
ICBS(TheIsraeliCentralBureauofStatistics)(2011).Unpublishedtimeseriesdataonthepriceofcerealsandexchangerate(2001–2010).
Kuiper,M.and.vanTongeren(2006).UsinggravitytomoveArmington‐anempiricalapproach to the small initial trade share problem in general equilibrium models.Paperpreparedfor theNinthAnnualConferenceonGlobalEconomicAnalysis, June15‐17,2006inAddisAbaba,Ethiopia(Version0.3,3May2006).
McDonald,S.(2009).Astaticappliedgeneralequilibriummodel(STAGE).Version1:July2007.DepartmentofEconomics&Strategy,OxfordBrookesUniversity,UK.
Mundlak, Y. and D. Larson (1992). On the transmission of world agriculture prices.WorldBankEconomicReview,6(3):399–422.
Nicita,A.(2006).MultilateralTradeLiberalizationandMexicanHouseholds:TheEffectof theDohaDevelopmentAgenda. In:Hertel,T.andL.A.Winters(eds.):PovertyandtheWTO:ImpactsoftheDohaDevelopmentAgenda.TheWorldBank:107‐128.
27
Sadoulet, E. and A. de Janvry (1995). Quantitative Development Policy Analysis. TheJohnsHopkinsUniversity Press, Baltimore and London, 1995.Manuscript availableat:Are.berkeley.edu/~sadoulet.
Siddig,K.,Flaig,D.,Luckman,J.andH.Grethe(2011).A2004SocialAccountingMatrixfor Israel. Agricultural Economics Working Paper Series (HohenheimerAgrarökonomischeArbeitsberichte),No.20.
28
8 Appendix:Descriptionoftheagriculturalsectorsofthe2004SAM
SAM2004sectors DetaileddescriptionbasedonIOT1995
Cereal:Growingofrice(non‐husked),corn,barley,rye,oats,andothercereals
Growingofothercerealsandpulses
Growingofwheat Growingofwheat
Fruitandvegetablesincludingmateandspices,excludingotherfruitdried
Growingofcitrus
Growingofpomefruits
Growingofstonefruits
Growingofnuts
Growingofgrapes
Growingofbananas
Growingofolives
Growingofsubtropicalandothertreecrops
Growingofvegetables(includingmelonsandpumpkins)
Growingofpotatoes
Othercrops:Growingofroughfodder,cotton,otherplantbasedfibers,oilseeds,soyabeans,groundnuts,tobacco.Growingofflowers,seeds,andnurseryproducts.Importsofgreencoffee,cocoabeans,andbulktea
Growingofroughfodder
Growingofcotton
Growingofotherfieldcrops(includinggroundnuts)
Bovinecattle,horses,asses,mules,andhinnies,live,bovinesemen,sheep,livegoatsshornwool,greasy,andfleece‐washedshornwool
Cattlefarmingformeat
Sheepfarming(milk,meat,andwool)
Otheranimal:Layinghens,poultryfarming,andturkeyfarming.Farmingofotheranimalsincludingfineanimalhair,notcardedorcombedandsilk‐wormcocoonssuitableforreeling
Raisingoflayinghens
Poultryfarmingformeat
Poultryhatcheries
Raisingofturkeysandotherpoultryfarming
Farmingofotheranimals
Meat:Processingofmeatandpoultry,includingCPCSubclass02962(pulledwool,greasy,includingfleece‐washedpulledwool;coarseanimalhair)andCPCSubclass02971(rawhidesandskinsofbovineorequineanimals,sheeporlambs,goatsorkids)
Processingofmeatandpoultry
Manufactureofedibleoils,margarine,andoilproductsexcludingCPCSubclass21660(maize(corn)oilanditsfractions,notchemicallymodified)
Manufactureofedibleoils,margarine,andoilproducts
29
SAM2004sectors DetaileddescriptionbasedonIOT1995
Milk:Milkofcattle,sheep,andgoats Cattlefarmingformilk
Forestry,fruittrees(investmenttobearing),andimportsofnaturalrubber
Forestry
Fruittrees(investmenttobearing)
Naturalrubber(imports)
Fishing:Pond‐culturefisheries.shoreandlakefisheries
Pond‐culturefisheries
Shoreandlakefisheries