7.2 Utility Maximization and Demand – Principles of Economics

Derive an individual demand curve from utility-maximizing adjustments to ... utility—that is, choices that follow the marginal decision rule—generally ... Skiptocontent LearningObjectives Deriveanindividualdemandcurvefromutility-maximizingadjustmentstochangesinprice. Derivethemarketdemandcurvefromthedemandcurvesofindividuals. Explainthesubstitutionandincomeeffectsofapricechange. Explaintheconceptsofnormalandinferiorgoodsintermsoftheincomeeffect. Choicesthatmaximizeutility—thatis,choicesthatfollowthemarginaldecisionrule—generallyproducedownward-slopingdemandcurves.Thissectionshowshowanindividual’sutility-maximizingchoicescanleadtoademandcurve. DerivinganIndividual’sDemandCurve Suppose,forsimplicity,thatMaryAndrewsconsumesonlyapples,denotedbytheletterA,andoranges,denotedbytheletterO.Applescost$2perpoundandorangescost$1perpound,andherbudgetallowshertospend$20permonthonthetwogoods.WeassumethatMs.Andrewswilladjustherconsumptionsothattheutility-maximizingconditionholdsforthetwogoods:Theratioofmarginalutilitytopriceisthesameforapplesandoranges.Thatis, Equation7.4 [latex]\frac{MU_A}{\$2}=\frac{MU_O}{\$1}[/latex] HereMUAandMUOarethemarginalutilitiesofapplesandoranges,respectively.Herspendingequalsherbudgetof$20permonth;supposeshebuys5poundsofapplesand10oforanges. Nowsupposethatanunusuallylargeharvestofappleslowerstheirpriceto$1perpound.Thelowerpriceofapplesincreasesthemarginalutilityofeach$1Ms.Andrewsspendsonapples,sothatathercurrentlevelofconsumptionofapplesandoranges Equation7.5 [latex]\frac{MU_A}{\$1}>\frac{MU_O}{\$1}[/latex] Ms.Andrewswillrespondbypurchasingmoreapples.Asshedoesso,themarginalutilityshereceivesfromappleswilldecline.Ifsheregardsapplesandorangesassubstitutes,shewillalsobuyfeweroranges.Thatwillcausethemarginalutilityoforangestorise.Shewillcontinuetoadjustherspendinguntilthemarginalutilityper$1spentisequalforbothgoods: Equation7.6 [latex]\frac{MU_A}{\$1}=\frac{MU_O}{\$1}[/latex] Supposethatatthisnewsolution,shepurchases12poundsofapplesand8poundsoforanges.Sheisstillspendingallofherbudgetof$20onthetwogoods[(12 x $1)+(8 x $1)=$20]. Figure7.3UtilityMaximizationandanIndividual’sDemandCurve MaryAndrews’sdemandcurveforapples,d,canbederivedbydeterminingthequantitiesofapplesshewillbuyateachprice.Thosequantitiesaredeterminedbytheapplicationofthemarginaldecisionruletoutilitymaximization.Atapriceof$2perpound,Ms.Andrewsmaximizesutilitybypurchasing5poundsofapplespermonth.Whenthepriceofapplesfallsto$1perpound,thequantityofapplesatwhichshemaximizesutilityincreasesto12poundspermonth. Itisthroughaconsumer’sreactiontodifferentpricesthatwetracetheconsumer’sdemandcurveforagood.Whenthepriceofappleswas$2perpound,Ms.Andrewsmaximizedherutilitybypurchasing5poundsofapples,asillustratedinFigure7.3“UtilityMaximizationandanIndividual’sDemandCurve”.Whenthepriceofapplesfell,sheincreasedthequantityofapplesshepurchasedto12pounds. HeadsUp! Noticethat,inthisexample,Ms.Andrewsmaximizesutilitywherenotonlytheratiosofmarginalutilitiestopriceareequal,butalsothemarginalutilitiesofbothgoodsareequal.But,theequal-marginal-utilityoutcomeisonlytrueherebecausethepricesofthetwogoodsarethesame:eachgoodispricedat$1inthiscase.Ifthepricesofapplesandorangesweredifferent,themarginalutilitiesattheutilitymaximizingsolutionwouldhavebeendifferent.Theconditionformaximizingutility—consumewheretheratiosofmarginalutilitytopriceareequal—holdsregardless.Theutility-maximizingconditionisnotthatconsumersmaximizeutilitybyequatingmarginalutilities. Figure7.4 Utilitymaximizingconditionis:[latex]\frac{MU_{X}}{P_{X}}=\frac{MU_{X}}{P_{Y}}[/latex] Utilitymaximizingconditionisnot:[latex]MU_{X}=MU_{Y}[/latex] FromIndividualtoMarketDemand ThemarketdemandcurveswestudiedinpreviouschaptersarederivedfromindividualdemandcurvessuchastheonedepictedinFigure7.3“UtilityMaximizationandanIndividual’sDemandCurve”.SupposethatinadditiontoMs.Andrews,therearetwootherconsumersinthemarketforapples—EllenSmithandKoyKeino.ThequantitieseachconsumesatvariouspricesaregiveninFigure7.5“DerivingaMarketDemandCurve”,alongwiththequantitiesthatMs.Andrewsconsumesateachprice.ThedemandcurvesforeachareshowninPanel(a).Themarketdemandcurveforallthreeconsumers,showninPanel(b),isthenfoundbyaddingthequantitiesdemandedateachpriceforallthreeconsumers.Atapriceof$2perpound,forexample,Ms.Andrewsdemands5poundsofapplespermonth,Ms.Smithdemands3pounds,andMr.Keinodemands8pounds.Atotalof16poundsofapplesaredemandedpermonthatthisprice.Addingtheindividualquantitiesdemandedat$1perpoundyieldsmarketdemandof40poundspermonth.Thismethodofaddingamountsalongthehorizontalaxisofagraphisreferredtoassumminghorizontally.Themarketdemandcurveisthusthehorizontalsummationofalltheindividualdemandcurves. Figure7.5DerivingaMarketDemandCurve ThedemandschedulesforMaryAndrews,EllenSmith,andKoyKeinoaregiveninthetable.TheirindividualdemandcurvesareplottedinPanel(a).ThemarketdemandcurveforallthreeisshowninPanel(b). Individualdemandcurves,then,reflectutility-maximizingadjustmentbyconsumerstovariousmarketprices.Onceagain,weseethatasthepricefalls,consumerstendtobuymoreofagood.Demandcurvesaredownward-slopingasthelawofdemandasserts. SubstitutionandIncomeEffects Wesawthatwhenthepriceofapplesfellfrom$2to$1perpound,MaryAndrewsincreasedthequantityofapplesshedemanded.Behindthatadjustment,however,lietwodistincteffects:thesubstitutioneffectandtheincomeeffect.Itisimportanttodistinguishtheseeffects,becausetheycanhavequitedifferentimplicationsfortheelasticityofthedemandcurve. First,thereductioninthepriceofapplesmadethemcheaperrelativetooranges.Beforethepricechange,itcostthesameamounttobuy2poundsoforangesor1poundofapples.Afterthepricechange,itcostthesameamounttobuy1poundofeitherorangesorapples.Ineffect,2poundsoforangeswouldexchangefor1poundofapplesbeforethepricechange,and1poundoforangeswouldexchangefor1poundofapplesafterthepricechange. Second,thepricereductionessentiallymadeconsumersofapplesricher.Beforethepricechange,Ms.Andrewswaspurchasing5poundsofapplesand10poundsoforangesatatotalcosttoherof$20.Atthenewlowerpriceofapples,shecouldpurchasethissamecombinationfor$15.Ineffect,thepricereductionforappleswasequivalenttohandinghera$5bill,therebyincreasingherpurchasingpower.Purchasingpowerreferstothequantityofgoodsandservicesthatcanbepurchasedwithagivenbudget. Todistinguishbetweenthesubstitutionandincomeeffects,economistsconsiderfirsttheimpactofapricechangewithnochangeintheconsumer’sabilitytopurchasegoodsandservices.Anincome-compensatedpricechangeAnimaginaryexerciseinwhichweassumethatwhenthepriceofagoodorservicechanges,theconsumersincomeisadjustedsothatheorshehasjustenoughtopurchasetheoriginalcombinationofgoodsandservicesatthenewsetofprices.isanimaginaryexerciseinwhichweassumethatwhenthepriceofagoodorservicechanges,theconsumer’sincomeisadjustedsothatheorshehasjustenoughtopurchasetheoriginalcombinationofgoodsandservicesatthenewsetofprices.Ms.Andrewswaspurchasing5poundsofapplesand10poundsoforangesbeforethepricechange.Buyingthatsamecombinationafterthepricechangewouldcost$15.Theincome-compensatedpricechangethusrequiresustotake$5fromMs.Andrewswhenthepriceofapplesfallsto$1perpound.Shecanstillbuy5poundsofapplesand10poundsoforanges.If,instead,thepriceofapplesincreased,wewouldgiveMs.Andrewsmoremoney(i.e.,wewould“compensate”her)sothatshecouldpurchasethesamecombinationofgoods. With$15andcheaperapples,Ms.Andrewscouldbuy5poundsofapplesand10poundsoforanges.Butwouldshe?Theanswerliesincomparingthemarginalbenefitofspendinganother$1onapplestothemarginalbenefitofspendinganother$1onoranges,asexpressedinEquation7.5.Itshowsthattheextrautilityper$1shecouldobtainfromapplesnowexceedstheextrautilityper$1fromoranges.Shewillthusincreaseherconsumptionofapples.Ifshehadonly$15,anyincreaseinherconsumptionofappleswouldrequireareductioninherconsumptionoforanges.Ineffect,sherespondstotheincome-compensatedpricechangeforapplesbysubstitutingapplesfororanges.Thechangeinaconsumer’sconsumptionofagoodinresponsetoanincome-compensatedpricechangeiscalledthesubstitutioneffectThechangeinaconsumersconsumptionofagoodinresponsetoanincome-compensatedpricechange.. Supposethatwithanincome-compensatedreductioninthepriceofapplesto$1perpound,Ms.Andrewswouldincreaseherconsumptionofapplesto9poundspermonthandreduceherconsumptionoforangesto6poundspermonth.Thesubstitutioneffectofthepricereductionisanincreaseinappleconsumptionof4poundspermonth. Thesubstitutioneffectalwaysinvolvesachangeinconsumptioninadirectionoppositethatofthepricechange.Whenaconsumerismaximizingutility,theratioofmarginalutilitytopriceisthesameforallgoods.Anincome-compensatedpricereductionincreasestheextrautilityperdollaravailablefromthegoodwhosepricehasfallen;aconsumerwillthuspurchasemoreofit.Anincome-compensatedpriceincreasereducestheextrautilityperdollarfromthegood;theconsumerwillpurchaselessofit. Inotherwords,whenthepriceofagoodfalls,peoplereacttothelowerpricebysubstitutingorswitchingtowardthatgood,buyingmoreofitandlessofothergoods,ifweartificiallyholdtheconsumer’sabilitytobuygoodsconstant.Whenthepriceofagoodgoesup,peoplereacttothehigherpricebysubstitutingorswitchingawayfromthatgood,buyinglessofitandinsteadbuyingmoreofothergoods.Byexaminingtheimpactofconsumerpurchasesofanincome-compensatedpricechange,wearelookingatjustthechangeinrelativepricesofgoodsandeliminatinganyimpactonconsumerbuyingthatcomesfromtheeffectivechangeintheconsumer’sabilitytopurchasegoodsandservices(thatis,weholdtheconsumer’spurchasingpowerconstant). Tocompleteouranalysisoftheimpactofthepricechange,wemustnowconsiderthe$5thatMs.Andrewseffectivelygainedfromit.Afterthepricereduction,itcostherjust$15tobuywhatcosther$20before.Shehas,ineffect,$5morethanshedidbefore.Heradditionalincomemayalsohaveaneffectonthenumberofapplessheconsumes.ThechangeinconsumptionofagoodresultingfromtheimplicitchangeinincomebecauseofapricechangeiscalledtheincomeeffectThechangeinconsumptionofagoodresultingfromtheimplicitchangeinincomebecauseofapricechange.ofapricechange.Whenthepriceofagoodrises,thereisanimplicitreductioninincome.Whenthepriceofagoodfalls,thereisanimplicitincrease.Whenthepriceofapplesfell,Ms.Andrews(whowasconsuming5poundsofapplespermonth)receivedanimplicitincreaseinincomeof$5. SupposeMs.Andrewsusesherimplicitincreaseinincometopurchase3morepoundsofapplesand2morepoundsoforangespermonth.Shehasalreadyincreasedherappleconsumptionto9poundspermonthbecauseofthesubstitutioneffect,sotheadded3poundsbringsherconsumptionlevelto12poundspermonth.Thatispreciselywhatweobservedwhenwederivedherdemandcurve;itisthechangewewouldobserveinthemarketplace.Weseenow,however,thatherincreaseinquantitydemandedconsistsofasubstitutioneffectandanincomeeffect.Figure7.6“TheSubstitutionandIncomeEffectsofaPriceChange”showsthecombinedeffectsofthepricechange. Figure7.6TheSubstitutionandIncomeEffectsofaPriceChange ThisdemandcurveforMs.AndrewswaspresentedinFigure7.5“DerivingaMarketDemandCurve”.Itshowsthatareductioninthepriceofapplesfrom$2to$1perpoundincreasesthequantityMs.Andrewsdemandsfrom5poundsofapplesto12.Thisgraphshowsthatthischangeconsistsofasubstitutioneffectandanincomeeffect.Thesubstitutioneffectincreasesthequantitydemandedby4pounds,theincomeeffectby3,foratotalincreaseinquantitydemandedof7pounds. Thesizeofthesubstitutioneffectdependsontherateatwhichthemarginalutilitiesofgoodschangeastheconsumeradjustsconsumptiontoapricechange.AsMs.Andrewsbuysmoreapplesandfeweroranges,themarginalutilityofappleswillfallandthemarginalutilityoforangeswillrise.Ifrelativelysmallchangesinquantitiesconsumedproducelargechangesinmarginalutilities,thesubstitutioneffectthatisrequiredtorestoretheequalityofmarginal-utility-to-priceratioswillbesmall.Ifmuchlargerchangesinquantitiesconsumedareneededtoproduceequivalentchangesinmarginalutilities,thenthesubstitutioneffectwillbelarge. Themagnitudeoftheincomeeffectofapricechangedependsonhowresponsivethedemandforagoodistoachangeinincomeandonhowimportantthegoodisinaconsumer’sbudget.Whenthepricechangesforagoodthatmakesupasubstantialfractionofaconsumer’sbudget,thechangeintheconsumer’sabilitytobuythingsissubstantial.Achangeinthepriceofagoodthatmakesupatrivialfractionofaconsumer’sbudget,however,haslittleeffectonhisorherpurchasingpower;theincomeeffectofsuchapricechangeissmall. Becauseeachconsumer’sresponsetoapricechangedependsonthesizesofthesubstitutionandincomeeffects,theseeffectsplayaroleindeterminingthepriceelasticityofdemand.Allotherthingsunchanged,thelargerthesubstitutioneffect,thegreatertheabsolutevalueofthepriceelasticityofdemand.Whentheincomeeffectmovesinthesamedirectionasthesubstitutioneffect,agreaterincomeeffectcontributestoagreaterpriceelasticityofdemandaswell.Thereare,however,casesinwhichthesubstitutionandincomeeffectsmoveinoppositedirections.Weshallexploretheseideasinthenextsection. NormalandInferiorGoods Thenatureoftheincomeeffectofapricechangedependsonwhetherthegoodisnormalorinferior.Theincomeeffectreinforcesthesubstitutioneffectinthecaseofnormalgoods;itworksintheoppositedirectionforinferiorgoods. NormalGoods Anormalgoodisonewhoseconsumptionincreaseswithanincreaseinincome.Whenthepriceofanormalgoodfalls,therearetwoidentifyingeffects: Thesubstitutioneffectcontributestoanincreaseinthequantitydemandedbecauseconsumerssubstitutemoreofthegoodforothergoods. Thereductioninpriceincreasestheconsumer’sabilitytobuygoods.Becausethegoodisnormal,thisincreaseinpurchasingpowerfurtherincreasesthequantityofthegooddemandedthroughtheincomeeffect. Inthecaseofanormalgood,then,thesubstitutionandincomeeffectsreinforceeachother.Ms.Andrews’sresponsetoapricereductionforapplesisatypicalresponsetoalowerpriceforanormalgood. Anincreaseinthepriceofanormalgoodworksinanequivalentfashion.Thehigherpricecausesconsumerstosubstitutemoreofothergoods,whosepricesarenowrelativelylower.Thesubstitutioneffectthusreducesthequantitydemanded.Thehigherpricealsoreducespurchasingpower,causingconsumerstoreduceconsumptionofthegoodviatheincomeeffect. InferiorGoods Inthechapterthatintroducedthemodelofdemandandsupply,wesawthataninferiorgoodisoneforwhichdemandfallswhenincomerises.Itislikelytobeagoodthatpeopledonotreallylikeverymuch.Whenincomesarelow,peopleconsumetheinferiorgoodbecauseitiswhattheycanafford.Astheirincomesriseandtheycanaffordsomethingtheylikebetter,theyconsumelessoftheinferiorgood.Whenthepriceofaninferiorgoodfalls,twothingshappen: Consumerswillsubstitutemoreoftheinferiorgoodforothergoodsbecauseitspricehasfallenrelativetothosegoods.Thequantitydemandedincreasesasaresultofthesubstitutioneffect. Thelowerpriceeffectivelymakesconsumersricher.But,becausethegoodisinferior,thisreducesquantitydemanded. Thecaseofinferiorgoodsisthusquitedifferentfromthatofnormalgoods.Theincomeeffectofapricechangeworksinadirectionoppositetothatofthesubstitutioneffectinthecaseofaninferiorgood,whereasitreinforcesthesubstitutioneffectinthecaseofanormalgood. Figure7.7SubstitutionandIncomeEffectsforInferiorGoods Thesubstitutionandincomeeffectsworkagainsteachotherinthecaseofinferiorgoods.TheconsumerbeginsatpointA,consumingq1unitsofthegoodatapriceP1.WhenthepricefallstoP2,theconsumermovestopointB,increasingquantitydemandedtoq2.Thesubstitutioneffectincreasesquantitydemandedtoqs,buttheincomeeffectreducesitfromqstoq2. Figure7.7“SubstitutionandIncomeEffectsforInferiorGoods”illustratesthesubstitutionandincomeeffectsofapricereductionforaninferiorgood.WhenthepricefallsfromP1toP2,thequantitydemandedbyaconsumerincreasesfromq1toq2.Thesubstitutioneffectincreasesquantitydemandedfromq1toqs.Buttheincomeeffectreducesquantitydemandedfromqstoq2;thesubstitutioneffectisstrongerthantheincomeeffect.Theresultisconsistentwiththelawofdemand:Areductioninpriceincreasesthequantitydemanded.Thequantitydemandedissmaller,however,thanitwouldbeifthegoodwerenormal.Inferiorgoodsarethereforelikelytohavelesselasticdemandthannormalgoods. KeyTakeaways Individualdemandcurvesreflectutility-maximizingadjustmentbyconsumerstochangesinprice. Marketdemandcurvesarefoundbysumminghorizontallythedemandcurvesofalltheconsumersinthemarket. Thesubstitutioneffectofapricechangechangesconsumptioninadirectionoppositetothepricechange. Theincomeeffectofapricechangereinforcesthesubstitutioneffectifthegoodisnormal;itmovesconsumptionintheoppositedirectionifthegoodisinferior. TryIt! IlanaDrakulichasanentertainmentbudgetof$200persemester,whichshedividesamongpurchasingCDs,goingtoconcerts,eatinginrestaurants,andsoforth.WhenthepriceofCDsfellfrom$20to$10,herpurchasesrosefrom5persemesterto10persemester.Whenaskedhowmanyshewouldhaveboughtifherbudgetconstraintwere$150(sincewith$150shecouldcontinuetobuy5CDsandasbeforestillhave$100forspendingonotheritems),shesaidshewouldhavebought8CDs.Whatisthesizeofhersubstitutioneffect?Herincomeeffect?AreCDsnormalorinferiorforher?Whichexhibit,Figure7.6“TheSubstitutionandIncomeEffectsofaPriceChange”orFigure7.7“SubstitutionandIncomeEffectsforInferiorGoods”,depictsmoreaccuratelyherdemandcurveforCDs? CaseinPoint:Found!AnUpward-SlopingDemandCurve Figure7.8 CharlesHaynes–rice–CCBY-SA2.0. Thefactthatincomeandsubstitutioneffectsmoveinoppositedirectionsinthecaseofinferiorgoodsraisesatantalizingpossibility:Whatiftheincomeeffectwerethestrongerofthetwo?Coulddemandcurvesbeupwardsloping? Theanswer,fromatheoreticalpointofview,isyes.IftheincomeeffectinFigure7.7“SubstitutionandIncomeEffectsforInferiorGoods”werelargerthanthesubstitutioneffect,thedecreaseinpricewouldreducethequantitydemandedbelowq1.Theresultwouldbeareductioninquantitydemandedinresponsetoareductioninprice.Thedemandcurvewouldbeupwardsloping! Thesuggestionthatagoodcouldhaveanupward-slopingdemandcurveisgenerallyattributedtoRobertGiffen,aBritishjournalistwhowrotewidelyoneconomicmatterslateinthenineteenthcentury.SuchgoodsarethuscalledGiffengoods.ToqualifyasaGiffengood,agoodmustbeinferiorandmusthaveanincomeeffectstrongenoughtoovercomethesubstitutioneffect.TheexampleoftencitedofapossibleGiffengoodisthepotatoduringtheIrishfamineof1845–1849.Empiricalanalysisbyeconomistsusingavailabledata,however,hasrefutedthenotionoftheupward-slopingdemandcurveforpotatoesatthattime.Themostconvincingpartsoftherefutationweretopointoutthat(a)giventhefamine,therewerenotmorepotatoesavailableforpurchasethenand(b)thepriceofpotatoesmaynothaveevenincreasedduringtheperiod! ArecentstudybyRobertJensenandNolanMiller,though,suggeststhepossiblediscoveryofapairofGiffengoods.TheybegantheirsearchbythinkingaboutthetypeofgoodthatwouldbelikelytoexhibitGiffenbehaviorandarguedthat,likepotatoesforthepoorIrish,itwouldbeamaindietarystapleofapoorpopulation.Insuchasituation,purchasesoftheitemaresuchalargepercentageofthedietofthepoorthatwhentheitem’spricerises,theimplicitincomeofthepoorfallsdrastically.Inordertosubsist,thepoorreduceconsumptionofothergoodssotheycanbuymoreofthestaple.Insodoing,theyareabletoreachacaloricintakethatishigherthanwhatcanbeachievedbybuyingmoreofotherpreferredfoodsthatunfortunatelysupplyfewercalories. TheirpreliminaryempiricalworkshowsthatinsouthernChinariceisaGiffengoodforpoorconsumerswhileinnorthernChinanoodlesareaGiffengood.Inbothcases,thebasicgood(riceornoodles)providescaloriesatarelativelylowcostanddominatesthediet,whilemeatisconsideredthetastierbuthighercost-per-caloriefood.Usingdetailedhouseholddata,theyestimatethatamongthepoorinsouthernChinaa10%increaseinthepriceofriceleadstoa10.4%increaseinriceconsumption.Forwealthierhouseholdsintheregion,riceisinferiorbutnotGiffen.Forbothgroupsofhouseholds,theincomeeffectofapricechangemovesconsumptionintheoppositedirectionofthesubstitutioneffect.Onlyinthepooresthouseholds,however,doesitswampthesubstitutioneffect,leadingtoanupward-slopingdemandcurveforriceforpoorhouseholds.InnorthernChina,theneteffectofapriceincreaseonquantitydemandedofnoodlesissmaller,thoughitstillleadstohighernoodleconsumptioninthepooresthouseholdsofthatregion. Inasimilarstudy,DavidMcKenzietestedwhethertortillaswereaGiffengoodforpoorMexicans.Hefound,however,thattheywereaninferiorgoodbutnotaGiffengood.HespeculatedthatthedifferentresultmaystemfrompoorMexicanshavingawiderrangeofsubstitutesavailabletothemthandothepoorinChina. BecausetheJensen/MillerstudyisthefirstvindicationoftheexistenceofaGiffengooddespiteaverylongsearch,theauthorshaveavoidedrushingtopublicationoftheirresults.Rather,theyhavemadeavailableapreliminaryversionofthestudyreportedonherewhilecontinuingtorefinetheirestimation. Sources:RobertJensenandNolanMiller,“GiffenBehavior:TheoryandEvidence,”KSGFacultyResearchWorkingPapersSeriesRWP02-014,2002availableatksghome.harvard.edu/~nmiller/giffen.htmlorhttp://ssrn.com/abstract=310863.Attheauthors’requestweincludethefollowingnoteonthepreliminaryversion:“Becausewehavereceivednumerousrequestsforthispaper,wearemakingthisearlydraftavailable.Theresultspresentedinthisversion,whilestronglysuggestiveofGiffenbehavior,arepreliminary.Inthenearfutureweexpecttoacquireadditionaldatathatwillallowustoreviseourestimationtechnique.Inparticular,monthlytemperature,precipitation,andotherweatherdatawillenableustouseaninstrumentalvariablesapproachtoaddressthepossibilitythattheobservedvariationinpricesisnotexogenous.Onceavailable,theinstrumentalvariablesresultswillbeincorporatedintofutureversionsofthepaper.”;DavidMcKenzie,“AreTortillasaGiffenGoodinMexico?”EconomicsBulletin15:1(2002):1–7. AnswertoTryIt!Problem OnehundredfiftydollarsistheincomethatallowsMs.Drakulictopurchasethesameitemsasbefore,andthuscanbeusedtomeasurethesubstitutioneffect.Lookingonlyattheincome-compensatedpricechange(thatis,holdinghertothesamepurchasingpowerasintheoriginalrelativepricesituation),wefindthatthesubstitutioneffectis3moreCDs(from5to8).TheCDsthatshebuysbeyond8constituteherincomeeffect;itis2CDs.Becausetheincomeeffectreinforcesthesubstitutioneffect,CDsareanormalgoodforherandherdemandcurveissimilartothatshowninFigure7.6“TheSubstitutionandIncomeEffectsofaPriceChange”. Previous/nextnavigation Previous:7.1TheConceptofUtility Next:7.3IndifferenceCurveAnalysis:AnAlternativeApproachtoUnderstandingConsumerChoice Backtotop License PrinciplesofEconomicsbyUniversityofMinnesotaislicensedunderaCreativeCommonsAttribution-NonCommercial-ShareAlike4.0InternationalLicense,exceptwhereotherwisenoted.