Three Essays on Product Differentiation and Strategy

MISSING IMAGE

Material Information

Title:
Three Essays on Product Differentiation and Strategy
Physical Description:
1 online resource (105 p.)
Language:
english
Creator:
Knight, David T
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Economics
Committee Chair:
HAMILTON,JONATHAN H
Committee Co-Chair:
BLAIR,ROGER D
Committee Members:
SLUTSKY,STEVEN M
MITRA,DEBANJAN

Subjects

Subjects / Keywords:
adversising -- bundling -- mergers
Economics -- Dissertations, Academic -- UF
Genre:
Economics thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The first chapter considers a vertically differentiated products market, in which consumers internalize the environmental harm caused by the products they consume. A brown incumbent falsely represents its product as more green than it actually is. Consumers' beliefs regarding the harm caused by consuming the incumbent's variant determine whether a greener entrant can profitably enter the market. Greenwash by the incumbent firm serves profitably to increase its own market share, but it may also have important implications for the market structure. I show that 1) consumers benefit from false advertising when entry is not deterred, and when consumers are able to correctly identify the incumbent's variant as the dirtier product; 2) reductions in the fixed cost of entry always benefit consumers; 3) entry always increases the aggregate amount of environmental harm; and 4) reductions in the fixed cost of entry always harm the environment. These results may inform market regulators and environmental groups that wish to combat (or facilitate) false advertising by polluting firms. The second chapter considers concert ticket bundling. Music concert tickets - as well as many other admission-based events - are often sold in bundles, and the existing literature does not adequately address the composition of an optimal bundle this setting. This chapter decomposes substitutability of component events into both the degree of horizontal differentiation between components and the additivity of consumers' preferences across them. It considers a monopoly setting in which two component events may be offered separately or as a pure bundle (i.e., mixed bundling is treated as infeasible). The chapter considers both the optimal pricing and bundling strategy when the degree of differentiation between the components is exogenously determined and the construction of the optimal bundle when the degree of differentiation may be selected by the monopolist. I find that, in terms of the degree of horizontal differentiation between the components, very similar and very different components are sold separately, whereas moderately differentiated components are offered as a bundle. I also demonstrate that capacity constraints affect the incentive to bundle - reducing (intensifying) it when the constraint applies to unbundled components separately (jointly). Moreover, capacity constraints decrease the level of differentiation between bundled components. Lastly, I find that duopolists offer more similar events separately than a monopolist would optimally offer as a bundle. The third chapter explores the effect of horizontal mergers on product quality, paying particular attention to the case of hospital combinations. I examine the role that market concentration and the proximity of merging hospitals plays in determining post-merger quality adjustments and welfare effects. Consumers are uniformly distributed along a Hotelling line, as are N hospitals. Initially, the N hospitals first select location, and then, set service quality and prices simultaneously. After two of the hospitals merge, all hospitals can adjust service quality and prices, but locations remain fixed at their pre-merger placements. I show that 1) mergers between two neighboring hospitals always reduce service quality; 2) mergers between two hospitals that are only separated by one hospital between them result in increased service quality and consumer welfare; 3) mergers between two hospitals that are separated by two or more hospitals between them do not increase the profits of merging hospitals; and 4) mergers between two hospitals that are separated by two of more hospitals between them do not affect service quality or consumer welfare.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by David T Knight.
Thesis:
Thesis (Ph.D.)--University of Florida, 2014.
Local:
Adviser: HAMILTON,JONATHAN H.
Local:
Co-adviser: BLAIR,ROGER D.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2014
System ID:
UFE0046672:00001


This item is only available as the following downloads:


Full Text

PAGE 1

THREEESSAYSONPRODUCTDIFFERENTIATIONANDSTRATEGYByDAVIDTHOMASKNIGHTJR.ADISSERTATIONPRESENTEDTOTHEGRADUATESCHOOLOFTHEUNIVERSITYOFFLORIDAINPARTIALFULFILLMENTOFTHEREQUIREMENTSFORTHEDEGREEOFDOCTOROFPHILOSOPHYUNIVERSITYOFFLORIDA2014

PAGE 2

c2014DavidThomasKnightJr. 2

PAGE 3

IdedicatethisdissertationtoMomandDad.Myworkistheirwork. 3

PAGE 4

ACKNOWLEDGMENTS MygoalofobtainingaPh.D.ineconomicshasgovernedmostofmyseriousdecisionsoverthepastelevenyears,bothprofessionalandpersonal.WhileIhavededicatedmuchenergyandefforttoarrivingatthepointofsubmittingaPh.D.dissertationandultimatelypreparingmyselfforacareerasanacademiceconomist,thisdocumentandtheeducationalpreparationthatitculminateswouldnotbepossiblewithoutthesupport,encouragement,andprofessionalguidanceofanumberofpeople.Iamthankfulforallthatmymentors,classmates,family,andfriendshavegivenme.IcannotlistallofthepeopletowhomIoweadebtofgratitude,butIwouldliketothankafewofthemindividually.Iwouldliketobeginbythankingmydoctoraladvisor,Dr.JonathanHamilton,whohassupportedmydevelopmentintoanacademiceconomistandprovidedinvaluableresearchadvicethroughoutthatprocess.ThisdissertationwouldnotbepossiblewithoutDr.Hamilton`swealthofknowledgeregardingtheeconomicsofproductdifferentiation.Otherresearchprojectshavealsobenetedimmenselyfromhisthoughtfuladvisingandcomments.Inadditiontoresearchsupport,Dr.HamiltonprovidedmetheexibilitythatallowedmetobecomethetypeofeconomistIwanttobe.Allowingmetopursueteachingdevelopmentasanimportantcomponentwithinmydoctoraleducationandtoseizeopportunitiesoutsideofthestandardcurriculum,hissupportandguidancearetheprimaryreasonsthatIamsowellpreparedtotransitionfromagraduatestudenttofacultymember.Iamveryfortunatetohavehadsuchanexcellentadvisor.IamalsogreatlyindebtedtoDr.RogerBlair,whohasprovidedmewithexcellentresearchopportunitiesandadditionalteachingsupport.MyrelationshipwithDr.BlairhaskeptmefullyengagedthroughoutthedissertationphaseofmyPh.D.andhisnudginghaskeptmeveryactiveinresearch.Iamveryluckytohavehadamentorthatcaressodeeplyabouthisstudentsandpushesthemtosucceedatthehighestlevel.IcanalwayscountonDr.Blairsthoughtful(andtruthful)advice. 4

PAGE 5

IwouldalsoliketothankDr.StevenSlutsky,whosededicationtotheeconomicsPh.D.programhasmadethepastveyearsanenjoyableexperience.Veryfewprogramsfosterthecollegialitythatoursdoes,andDr.Slutskyisresponsibleforfacilitatingthatenvironment.Hisinsistencethatourprogramisoneofacollaborative,asopposedtocompetitive,characterallowedmyclassmatesandmetheopportunitytoshareideasandlearnfromeachother.Whilehehasnotbeendirectlyinvolvedintheresearchcomponentofmyeducation,IwouldnotbetheeconomistIamwithoutDr.MarkRush.HeisresponsibleformycontinueddevelopmentasaneducatorandthereasonthatIhaveachievedtheteachingsuccessesIhave.Hisadviceregardinginstructionandstudentinteractionshasbeeninvaluable,andhisfriendshipevenmoreinvaluable.IwouldalsoliketothankDr.CatherineElliottforhersupportofmeasanundergraduateeconomicsstudent.Dr.ElliottisthereasonthatIdecidedtopursuegraduateeducationineconomicsandthereasonIwantedtobecomeaneconomicsfacultymember.Herimpressivededicationtoandexceptionalperformanceintheclassroomhaveprovidedmewitharolemodel.Iamthankfulforherearlyinuenceovermyeducation,andIwillcontinuetoaskmyself,WhatwouldElliottdo?Inadditiontothefacultymembersthathavebeensoinuentialoverthepastelevenyears,Iwouldliketothankmyclassmates,fromNewCollegeofFloridatoTilburgUniversitytotheUniversityofFlorida.Withoutthesefriendsandcolleagues,Iwouldhavenevermadeittothispoint.Theirassistancewithcompletinghomeworkassignmentsandstudyingforexamswasinstrumentalinmemakingitthisfar,andwithouttheirupliftingspirits,itneverwouldhavebeensuchanenjoyableelevenyears.Again,Icannotthankeveryone,butIoweaspecialdebtofgratitudetoLizHudson,KarlyeDilts,AntonioLourenco,DavidBrown,BryanLynch,andLindseyWoodworth.Lastly,Iwouldliketothankmyfamilyfortheirloveandsupportoverthelastthirtyyears.VeryfewpeoplearesofortunateastohavethesupportivefamilyIenjoy.Ithank 5

PAGE 6

mymomforthecountlessnightsofhomeworkhelpandconstantemotionalsupport,mydadforallowingmetopersiston(andonandon),mysistersforbeingmyownpersonalcheerleadingsquad,andGerardoforgroundingmeandshowingmewhatIreallywantoutofallofthis. 6

PAGE 7

TABLEOFCONTENTS page ACKNOWLEDGMENTS .................................. 4 LISTOFFIGURES ..................................... 8 ABSTRACT ......................................... 9 CHAPTER 1DETERRINGTHEGREENREVOLUTION:GREENWASHANDTHETHREATOFGREENENTRY ................................. 11 1.1Introduction ................................... 11 1.2PreviousLiterature ............................... 13 1.3TheoreticalModel:VerticalDifferentiation .................. 17 1.3.1Stages2and3:PriceSettingandtheEffectonGreenEntry .... 20 1.3.2Stage1:Greenwashing ........................ 28 1.4ImplicationsforConsumerWelfareandEnvironmentalHarm ....... 43 1.5ConcludingRemarks .............................. 48 2SPEARSANDDYLAN?CHARACTERIZINGTHEINCENTIVETOBUNDLEWHENMIXEDBUNDLINGISNOTFEASIBLE .................. 51 2.1Introduction ................................... 51 2.2ExtantLiterature ................................ 53 2.3TheModel .................................... 54 2.4Extension:CapacityConstraints ....................... 70 2.5Conclusion ................................... 75 3HOSPITALMERGERSANDPRODUCTDIFFERENTIATION:ANECONOMICAPPROACH ..................................... 78 3.1Introduction ................................... 78 3.2HealthcareCombinationsandServiceQuality ................ 80 3.3TheoreticalModelofVerticalandHorizontalDifferentiation ........ 82 3.4Conclusion ................................... 91 APPENDIX:MATHEMATICALDERIVATIONS ...................... 93 REFERENCES ....................................... 101 BIOGRAPHICALSKETCH ................................ 105 7

PAGE 8

LISTOFFIGURES Figure page 1-1TheEffectofkonReportedHarm ......................... 41 1-2TheEffectofkonConsumerSurplus ....................... 45 1-3TheEffectofkonAggregateEnvironmentalHarm ................ 47 2-1Multi-purchasing ................................... 59 2-2ComponentPricing .................................. 62 2-3BundleDemand ................................... 64 2-4TheIncentivetoBundle ............................... 67 2-5CapacityConstraintsImposedOnUnbundledComponentsSeparately ..... 73 2-6CapacityConstraintsImposedOnUnbundledComponentsJointly ....... 74 2-7CapacityConstraintsCompared .......................... 75 3-1PatientChoice .................................... 84 8

PAGE 9

AbstractofDissertationPresentedtotheGraduateSchooloftheUniversityofFloridainPartialFulllmentoftheRequirementsfortheDegreeofDoctorofPhilosophyTHREEESSAYSONPRODUCTDIFFERENTIATIONANDSTRATEGYByDavidThomasKnightJr.May2014Chair:JonathanHamiltonMajor:EconomicsTherstchapterconsidersaverticallydifferentiatedproductsmarket,inwhichconsumersinternalizetheenvironmentalharmcausedbytheproductstheyconsume.Abrownincumbentfalselyrepresentsitsproductasmoregreenthanitactuallyis.Consumers'beliefsregardingtheharmcausedbyconsumingtheincumbentsvariantdeterminewhetheragreenerentrantcanprotablyenterthemarket.Greenwashbytheincumbentrmservesprotablytoincreaseitsownmarketshare,butitmayalsohaveimportantimplicationsforthemarketstructure.Ishowthat1)consumersbenetfromfalseadvertisingwhenentryisnotdeterred,andwhenconsumersareabletocorrectlyidentifytheincumbentsvariantasthedirtierproduct;2)reductionsinthexedcostofentryalwaysbenetconsumers;3)entryalwaysincreasestheaggregateamountofenvironmentalharm;and4)reductionsinthexedcostofentryalwaysharmtheenvironment.Theseresultsmayinformmarketregulatorsandenvironmentalgroupsthatwishtocombat(orfacilitate)falseadvertisingbypollutingrms.Thesecondchapterconsidersconcertticketbundling.Musicconcertticketsaswellasmanyotheradmission-basedeventsareoftensoldinbundles,andtheexistingliteraturedoesnotadequatelyaddressthecompositionofanoptimalbundlethissetting.Thischapterdecomposessubstitutabilityofcomponenteventsintoboththedegreeofhorizontaldifferentiationbetweencomponentsandtheadditivityofconsumerspreferencesacrossthem.Itconsidersamonopolysettinginwhichtwo 9

PAGE 10

componenteventsmaybeofferedseparatelyorasapurebundle(i.e.,mixedbundlingistreatedasinfeasible).Thechapterconsidersboththeoptimalpricingandbundlingstrategywhenthedegreeofdifferentiationbetweenthecomponentsisexogenouslydeterminedandtheconstructionoftheoptimalbundlewhenthedegreeofdifferentiationmaybeselectedbythemonopolist.Indthat,intermsofthedegreeofhorizontaldifferentiationbetweenthecomponents,verysimilarandverydifferentcomponentsaresoldseparately,whereasmoderatelydifferentiatedcomponentsareofferedasabundle.Ialsodemonstratethatcapacityconstraintsaffecttheincentivetobundlereducing(intensifying)itwhentheconstraintappliestounbundledcomponentsseparately(jointly).Moreover,capacityconstraintsdecreasethelevelofdifferentiationbetweenbundledcomponents.Lastly,Indthatduopolistsoffermoresimilareventsseparatelythanamonopolistwouldoptimallyofferasabundle.Thethirdchapterexplorestheeffectofhorizontalmergersonproductquality,payingparticularattentiontothecaseofhospitalcombinations.Iexaminetherolethatmarketconcentrationandtheproximityofmerginghospitalsplaysindeterminingpost-mergerqualityadjustmentsandwelfareeffects.ConsumersareuniformlydistributedalongaHotellingline,asareNhospitals.Initially,theNhospitalsrstselectlocation,andthen,setservicequalityandpricessimultaneously.Aftertwoofthehospitalsmerge,allhospitalscanadjustservicequalityandprices,butlocationsremainxedattheirpre-mergerplacements.Ishowthat1)mergersbetweentwoneighboringhospitalsalwaysreduceservicequality;2)mergersbetweentwohospitalsthatareonlyseparatedbyonehospitalbetweenthemresultinincreasedservicequalityandconsumerwelfare;3)mergersbetweentwohospitalsthatareseparatedbytwoormorehospitalsbetweenthemdonotincreasetheprotsofmerginghospitals;and4)mergersbetweentwohospitalsthatareseparatedbytwoofmorehospitalsbetweenthemdonotaffectservicequalityorconsumerwelfare. 10

PAGE 11

CHAPTER1DETERRINGTHEGREENREVOLUTION:GREENWASHANDTHETHREATOFGREENENTRY 1.1IntroductionClaimsofvariousproducts'environmentalfriendlinesshavebecomecommonplace.Organicproduce,shade-growncoffee,andlead-freetoysofferfamiliarexamplesofattemptsbyrmstoconveyenvironmentalfriendlinesstoconsumers.Notsurprisingly,theriseingreen-basedadvertisingaccompaniesamarkedriseinconsumers`willingnesstopayforenvironmentallyfriendlyproducts.Boththeprovisionofinformationbyrmsandtheinternalizationofenvironmentalexternalitiesbyconsumersshouldenhancemarketefciencyandreduceenvironmentalharm,buttheyalsogiverisetonewissuesofconcern.Dormshavetheincentivetomakehonestclaimsabouttheirproducts,andaretheliestheymighttellnecessarilybad?Falseclaimsofenvironmentalfriendlinessorstewardship-commonlyreferredtoasgreenwash-misleadconsumersanddupethemintopayingelevatedpricesforbrownproducts,orevenbuyingbrownproductsthattheyotherwisewouldnothavepurchased.Greenwashisastrategictoolemployedbyrmstomisleadconsumersandcheaplyincreasethedemandfortheirproducts.Deceitfullyincreasingthedemandforaproductthroughfalseclaimsofelevatedqualityallowsrmstochargehigherpricesandearnsupranormalprotswithoutactuallymakingcostlyqualityimprovements.Reducingfalseenvironmentalclaimsshould,atrstglance,beconsideredsociallydesirable.Forthesereasons,theFTCseekstolimitsuchdeceptivepractices.Section5oftheFederalTradeCommissionAct(FTCAct)prohibitsunfairordeceptiveactsorpracticesinoraffectingcommerceandmakestheFederalTradeCommission(FTC)responsibleformonitoringandpunishingfalseadvertisingclaims.Inresponsetoasharpriseinenvironment-relatedclaims,lastyear,theFTCissuedGuidesfortheUseofEnvironmentalMarketingClaims.Thesecomprehensiveguidelinesprovidetheagency`sinterpretationofSection5oftheFTCAct,asitapplies 11

PAGE 12

togreenadvertising.Theyofferaglimpseatwhatpracticestheagencyislikelytochallenge.Whilethesethoroughguidelinesadvocateforscienticallyveriableandspecicclaims,theycannotbeconsideredexhaustive.Theysimplyaddressthemostcommonpracticesingreenadvertising;morenuancedcasesaretobeaddressedbythecommissionatitsowndiscretion.Astheagencymovesforwardanddecideswhatcasestotakeup,itshouldbeinformedoftheincentivesthatitsbehavioraffectsandthelikelyoutcomesofghtingfalsegreenadvertisingclaims.Thispaperexplorestheeffectsofenvironmentalmarketingandhighlightsanumberofimportantissuesthatshouldbeconsideredbyanenforcementagency.Itcharacterizestheeffectofgreenwashadvertisingonconsumersurplusandaggregateenvironmentalharm.Thefollowingsectionspresentandanalyzeamulti-stagegame,inwhichtwormseachproduceonevariantofadifferentiatedproduct,eachindexedbytheexogenouslyassignedlevelofenvironmentalharmityields.Abrownincumbentproducesavariantthatisharmfultotheenvironment,andagreenpotentialentrantproducesalessharmfulvariant.Beforethegreenrmdecidestoenter,thebrownincumbentmayundertakefalseadvertisingtoconvinceconsumersthatitsvariantislessharmfulthatitactuallyis.1Theremainderofthispaperisorganizedasfollows:Section2proceedswithabriefsurveyoftherecentliteratureongreenwashandgreenentry,Section3presentsandsolvesatheoreticalmodelofgreenwashbyabrownincumbentwhenitfacesthethreatofentrybyagreenerrival,andnally,Section4presentstheeffectsofgreenwashonconsumerwelfareandaggregateenvironmentalharm,andSection5providessomebriefconcludingremarks. 1Themodelpresentedinthispapertreatsthepollutingproductasacredencegood,assumingthatneitherconsumersnorthegreenpotentialentrantcanobservetheincumbent`sactuallevelofharmexpost. 12

PAGE 13

1.2PreviousLiteratureIncreasingly,economicinquiriesrelatedtopollutingproductstreattheproducts'environmentaleffectsasdimensionsofverticaldifferentiation,asopposedtothetraditionalexternalities-basedtreatment.Thisrelativelynewviewofenvironmentalharmpresentsnewissuesrelatedtostrategicbehavior,asenvironmentalharmbecomesincorporatedintothepayoffsandincentivesofeconomicagents.Thispaperconsiderstherolethatfalseadvertisingplaysindeterminingmarketstructureandenvironmentalharm.Thissectionplacesthispaperincontext,bypresentingabriefreviewoftwostringsofcurrentresearch.First,thissectionidentiesanewapproachtoanalyzingpollutingrmsandtheirproducts,onebasedonverticaldifferentiation.Thissectionalsoidentiesanemergingliteraturerelatedtofalseadvertisinganditseffectonmarketefciency,payingparticularattentiontopollutingproducts.2Below,eachofthesetwostringsofliteratureisbroughttogether,inanattempttoplacethispaperappropriatelyincontextanddemonstrateitscontribution.Therststringofrelevantliteraturedevelopstheverticaldifferentiationapproachtopollutingproducts.Manyrecenttheoreticalstudiesofpollutingproductsandenvironmentalharmutilizeamodelofverticallydifferentiatedproducts,originallyintroducedbyAroraandGangopadhyay(1995)andfurtherdevelopedbyCremerandThisse(1999)andMottaandThisse(1999).Forexample,Heijen(2007)employsaverticaldifferentiationapproachsimilartothatinCremerandThisse(1999).Heexaminestheroleofenvironmentalgroupsthatconductinformativeadvertisingcampaigns.Hendsthatwhenrmscannotchoosetheirlevelofharm,theenvironmentalgroupadvertisesagainstdirtyrms.Whenrmscanchoosetheirlevelofenvironmentalharm,simplythethreatoftheenvironmentalgroupisenoughtomakedirtyrmsreducetheirenvironmentalharm.HeijenandSchoonbeek(2008)obtainasimilarresultto 2Foranextensivesurveyoftheadvertisingliterature,seeBagwell(2007). 13

PAGE 14

thatreportedbyHeijen(2007).Theyassumethatconsumerscorrectlyobservetheenvironmentalharmrenderedbydifferentproducts.Insteadofinformingconsumers,theenvironmentalgroupundertakesacostlyadvertisingcampaignthatincreasestheconsumers'distasteforenvironmentalharm.Regardlessofwhetherarmcanaffectitslevelofenvironmentalharm,thepresenceofanenvironmentalgroupreducestheaggregatelevelofenvironmentalharm.Thecurrentpaperassumesthatconsumers'preferencescannotbealtered,butconsumerscannotperfectlyobservetheenvironmentalharmrenderedbytheproductstheyconsume.Anincumbentrmmisreportsitsharmthroughfalseadvertising.TheanalysispresentedinthispaperstandsinstarkcontrastfromHeijenandSchoonbeek(2008),inthatalladvertisingisconductedbytherm.Thereisnomarketregulatororenvironmentalgroup.VanderMadeandSchoonbeek(2009)examinetheeffectofconsumers'preferencesforgreenproductsonentrybygreenrms.SimilartotheanalysisinHeijenandSchoonbeek(2008),vanderMadeandSchoonbeek(2009)considerconsumersthatcorrectlyobservetheamountofenvironmentalharmcausedbythegoodstheyconsume,butwhosepreferencesmaybeaffectedbyanenvironmentalgroup`sadvertisingcampaign.Here,consumers'information-nottheirpreferences-determinethepotentialforprotableentrybyagreenrmandtheoverallmarketstructure.Asecondstringofrelevantliteratureexaminestheinteractionbetweenfalseadvertisingandmarketoutcomes.Theabilityofadvertisingtoincreasedemandhaslongbeenacknowledgedbyeconomists.3Insomeinstances,thismayenhancemarketefciency.Forexample,advertisingcanreducesearchcostsandsolveotherinformationimperfections,resultinginmoreefcientmarkets.However,itisalsoplausiblethatadvertisingmisinformsconsumersordistortstheirpreferences,potentiallyreducingefciency.Inthecontextofadvertisingforpollutingproducts,misinformationand 3See,forexample,DixitandNorman(1978),Salop(1979),andSchmalensee(1982). 14

PAGE 15

distortedpreferenceshavebecomeanimportantfocusofstudy.Anumberofrecentstudieshaveexaminedthewelfareeffectsoffalseadvertising,payingparticularattentiontomarketsforpollutingproducts.GlaeserandUjhelyi(2010)considerfalseadvertisinginaCournotoligopolysettingandderivethewelfareeffectsfortheunregulatedandregulatedequilibria.Theydemonstratethatfalseadvertisingmaybewelfareenhancingwhenundertakenbyamonopolythatcannotchangeitspost-advertisingprice.Inthislimitedsetting,falseadvertisingmayenhancewelfarebyoffsettingthedeadweightlossassociatedwithmonopolypricing.Relatedly,theydemonstratethatmisinformationalwaysreducesconsumersurplus.Thepresentpaperdoesnotconsidergovernmentpoliciesaimedacurbingfalseadvertising,butitdoesconsiderthewelfareeffectoffalseadvertising,ina(vertically)differentiatedproductssetting.HattoriandHigashida(2011)andHattoriandHigashida(2012)considertheeffectsoffalseadvertisingindifferentiatedproductsmarkets.HattoriandHigashida(2011)examinestheeffectsoffalseadvertisinginamarketcharacterizedbyhorizontallydifferentiatedproducts.Inthissetting,theyhighlighttheimportantrolethatanadvertisingexternalityplaysindeterminingtheeffectoffalseadvertisingonwelfare.Whentheadvertisingspilloversarelarge,andgoodsaremorehomogenous,lessadvertisingoccursandconsumers'switchdecisionsarelessharmfultowelfare.However,HattoriandHigashida(2011)donotidentifycasesinwhichfalseadvertisinginahorizontallydifferentiatedproductsmarketmayenhanceconsumersurplusorwelfare.HattoriandHigashida(2012)conductasimilaranalysisinthecontextofaverticallydifferentiatedproductmarket.Theydemonstratethatthedominanteffectoffalseadvertisingonconsumersurplusarisesfrompricecompetition.Theauthorsdemonstratethatfalseadvertisingbythelowqualityrmmayincreaseconsumersurplus,throughmorevigorouspricecompetition.Conversely,falseadvertisingbythehighqualityformalwaysreduceswelfare,asitsoftenspricecompetition.Asimilarwelfareresultisobtainedinthefollowingsectionsofthecurrentpaper. 15

PAGE 16

Thissectionidentiestherecenttreatmentofenvironmentalharmasafactorintroducingverticalproductdifferentiation.Itidentiesarecentliteraturerelatedtotheeffectsofinformativeadvertising(byenvironmentalgroups)inpollutingindustries.Informativeadvertisingthataffectsconsumers'distasteforpollutionmayincentivizepollutionabatementandfacilitateentrybygreenrms.Thispaperdiffersfrompreviousstudiesofgreenadvertising.Thecurrentpaperassumesthatconsumers'preferencesareimmutable,andthatgreenadvertisingdoesnotinformconsumersofactualquality.Thecurrentpaperonlyconsidersgreenwash.Thissectionalsoidentiesanacceptedroleoffalseadvertisingasastrategicinstrument,availabletopollutingrmswhentheconsumerstheyserveinternalizeenvironmentalharm.Itpresentsanumberofresultsrelatedtothewelfareeffectsoffalseadvertisinganddemonstratesthatthewelfareeffectsaremarket-specic.Thatis,falseadvertisinghasambiguouseffectsonconsumersurplusandwelfare,thatcannotberesolvedwithoutdetailedinformationonmarketstructureandconduct.Thecurrentpaperconsidersmarketentry,andthusdealswithchangingmarketstructures.Asthestructureofthemarketchanges,thewelfareandenvironmentalimplicationsofgreenwasharenotstraightforward.Thispapercombinesbothofthesestringsofliterature,asitconsidersstrategicfalseadvertisinginaverticallydifferentiatedmarketforapollutingproduct.Thecurrentpapermakesanumberofimportantcontributionstotheextantliterature.Itistherstpaper,tomyknowledge,thatconsidersgreenwashadvertisingasaninstrumentcapableofalteringmarketstructure.Previousstudiesthatanalyzegreenwashadvertisinghaveconsideredmanagerialincentivesorconsumerbeliefsinthepresenceofgreenwash(See,forexample,LyonandMaxwell(2011)andMason(2013));nonehasconsideredmarket-leveleffects.Additionally,theresultspresentedinthispaperaresurprisingandinformative,astheyprovideanargumentinfavorofsomepositivelevelofgreenwash.Thepaperdemonstratesthatfalseadvertisingcanenhanceconsumerwelfareandreducepollution. 16

PAGE 17

Section3ofthispaperanalyzesthecompetitiveeffectsofgreenadvertisingbyanincumbentmonopolistinaverticallyorientedmarket,whereproductsaredifferentiatedbytheamountofenvironmentalharmresultingfromtheirproduction.Theanalysisdemonstratestheabilityofanincumbentmonopolisttostrategicallyemploygreenadvertisingtoprevententrybyagreenerentrant.Moreover,itexaminestheeffectofthisstrategicbehavioronmarketstructure.Asidentiedintheextantliterature,theeffectsoffalseadvertisingonwelfareandenvironmentalharmsubstantivelychangeasthemarketstructureisaltered.Section4presentsawelfareanalysisandtheimplicationsforaggregateenvironmentalharm.Section5providesbriefconcludingremarksanddiscussessomeareasforfutureresearch. 1.3TheoreticalModel:VerticalDifferentiationThispaperemploysagametheoreticmodelofverticallydifferentiatedproducts.Productsdifferintheexogenouslyassignedlevelofenvironmentalharmthattheirproductioncreates,butareotherwiseidentical.Thismodelisappliedtoamarketservedbyanincumbentmonopolistandagreenerpotentialentrant.Beforethegreenrmmakesitsentrydecision,thebrownincumbentmayundertakeacostlygreenwashingcampaign,makingitsproductappearmoregreenthanitactuallyis.Theanalysispresentedinthissectionpaysparticularattentiontostrategicentrydeterringbehaviorbytheincumbentmonopolistanditseffectonconsumerwelfareandaggregateenvironmentalharm.Themodelassumesthatacontinuumofotherwise-homogenousconsumersdifferintheir(dis)tasteforenvironmentalharm.Eachconsumerinternalizesaportionoftheenvironmentalharmrenderedbytheproducttheyconsume.Basedontheirmarginalvaluationsofthepollutionattribute,consumersdecidewhethertopurchaseoneunitoftheavailablevariants,ortoabstainfrompurchasingthepollutingproductaltogether.Someconsumers'(dis)tastesforpollutionaresufcientlyhightoensurethattheydonotconsumetheproduct.Nofullcoverageassumptionismade;theproducerofthe 17

PAGE 18

highestqualityavailablevariantcanaffectthetotalnumberofconsumersthatchoosetopurchasetheproductorthetotalquantitytransacted.4Thispaperconsiderstwopossiblevariantsofthepollutingproduct.Thevariantsdifferinthe(exogenouslyassigned)levelofpollutionthateachcreates;thegreen(er)variantisassumedtobelessharmfulthantheincumbentmonopolist'svariant.Consumersperfectlyobservetheamountofharmcreatedbythenon-advertisedgreenproduct,buttheirperceptionofthebrownvariant'sharmmaybealteredbyenvironment-relatedadvertising,orgreenwashing.Aconsumerwith(dis)tasteparameterderivesthefollowingutilityfrompurchasingoneunitoftheithvariantofthedifferentiatedproduct: U()=V)]TJ /F6 11.955 Tf 11.95 0 Td[(ei)]TJ /F7 11.955 Tf 11.96 0 Td[(pi,(1)whereV>0isthegrossvaluationoftheproduct.eiandpiaretheenvironmentalharmandmarketpriceassociatedwithconsumingoneunitoftheithrm'svariant,respectively.Theutilityderivedfromnotpurchasingeithervariantoftheproductisnormalizedtozero.Itisassumedforsimplicitythattastesforenvironmentalharm,,areuniformlydistributedontheinterval[0,1].Consumerspurchaseoneunitofthevariantthatoffersthehighestlevelofutility.Absentadvertising,someindividualshaveverystrongpreferencesforgreenproductsandchoosetoabstainfromconsumingthepollutingproduct.Oftheindividualsthatdopurchaseoneofthetwoavailablevariants,consumersthathaveastrongerpreferenceforgreenproducts(i.e.,ahighervalueof)payahigherpriceforthegreenvariant,whileconsumerswithaweakerpreferenceforenvironmentalcleanlinesspurchasethebrownvariantatalowerprice.Fully-informedconsumersmakeefcientconsumptionchoicesbyinternalizingenvironmentalharm. 4Fullcoveragemaybeachievedwithanextremelystronglevelofgreenwash. 18

PAGE 19

Consumersmaynotcorrectlyobservetheenvironmentalharmdifferencesoftheproductstheyconsume,asacknowledgedbyanumberofanalysesofdifferentiatedproducts.5Instead,theymayrelyoninformationprovidedtothembythepollutingrmsand/orthird-partyagents(e.g.,environmentaladvocacygroups).Here,itisassumedthatconsumers'onlysourceofinformationisthermproducingthevariant.Thebrownrmmaymisleadconsumersbyundertakingacostlygreenwashingcampaignandreducingthelevelofperceivedenvironmentalharm.Thebrownrmincursapositiveadvertisingcostifitwishestoadvertiseitsproductasbeingmoregreenthanitactuallyis.Suchdeceitfulbehaviorservestodistortthemarketdemandinfavorofthebrownrm.6Thispaperdoesnotconsidergreenwashingbythegreenrm,asthetimingstructureofthegameallowsforprecommittmentbytheincumbent.Firmsdonotincuranymarginalproductioncosts.7Rather,theyincuraxedentryfee,F,whichtheincumbentisassumedtohavepaidbeforethestartofthegame.Whentheincumbentengagesingreenwashing,thecostoffalselyreportingareducedlevelofenvironmentalharm,^eB,toconsumersisgivenasc(eB)]TJ /F8 11.955 Tf 15.55 0 Td[(^eB)=k(eB)]TJ /F10 7.97 Tf 9.01 0 Td[(^eB)2 2,whereeBistheactualenvironmentalharmassociatedwiththeincumbent'svariant.Theparameterkmeasuresthehowcostlyitistomisleadconsumers.8Importantly,theactuallevels 5Forexample,seeHeijenandSchoonbeek(2008),vanderMadeandSchoonbeek(2009),GlaerserandUjhelyi(2010),HattoriandHigashida(2011a),andHattoriandHigashida(2011b).6Thispaperdoesnotconsidergreenadvertisingthatservestoinformconsumersofactualproductdifferences.Instead,itassumesthatrmscancostlesslyrevealtheiractualenvironmentalharm.7Includingmarginalproductioncostsisquitesimple,buttheydonotenhancetheanalysispresentedinthepresentpaper.8AsnotedbyHattoriandHigashida(2011)andHattoriandHigashida(2012),thecostoffalseadvertisingcapturesthegullibilityofconsumers.Thatis,thecostofmisleadingconsumersisdeterminedbytheirnaivety. 19

PAGE 20

ofenvironmentalharmareexogenouslyassignedanddonotchange(i.e.,eBandeGarexed.).Thecostofmisrepresentingenvironmentalharmisincreasinginthelevelofmisrepresentationatanincreasingrate.Thisconvexcoststructureensuresthattheincumbenthasaunique,nonzeroprot-maximizingreportedlevelofharm.Additionally,itdoesnotpreventtheincumbentfromattemptingtoconveyitsvariantasmoregreenthantheentrant'svariant.Thissectionproceedswithananalysisofthethree-stagegamedescribedabove.Theincumbentrmundertakessome(positive)amountofgreenwashadvertising,thegreenentrantchooseswhethertoenter,andtheextantrmssetprices.Theissueofprimaryconcerniswhethertheincumbentundertakesalevelofgreenwashthataltersthemarketstructure,andhowsuchbehavioraffectsconsumerwelfareandaggregateenvironmentalharm.InordertoidentifycandidateSubgamePerfectNashEquilibria,thissectionisorganizedaccordingtobackwardinduction,beginningwiththepricesettingandentrystages. 1.3.1Stages2and3:PriceSettingandtheEffectonGreenEntryAbrownincumbentrmgreenwashesitsproductwithoutthepossibilityofbeingexposedbytheenvironmentalgroup.Itfalselyadvertisesitsproductbyreportingalevelofenvironmentalharmthatislessthantheactuallevel.Intheabsenceofagovernmentregulatororanenvironmentalgroup,consumersaccepttherm'sreportofitsenvironmentalharmandmaketheirpurchasechoicesaccordingly.Thenalstageofthegameischaracterizedbypricecompetition,inwhicheachrmsetsapriceforitsvariant.InordertofullycharacterizeallpossibleSubgamePerfectNashEquilibria,wemustconsiderfourmarketarrangementsthatcouldariseinthenalstageofthegame:1)Thebrownvariantisreportedasbeinglessenvironmentallyfriendlythanthegreenvariant,andthegreenrmenters;2)Thebrownvariantisreportedasbeinglessenvironmentallyfriendlythanthegreenvariant,andthegreenrmdoesnotenter;3)Thebrownvariantisreportedasbeingmoreenvironmentallyfriendlythanthegreen 20

PAGE 21

variant,andthegreenrmdoesnotenter;and4)Thebrownvariantisreportedasbeingmoreenvironmentallyfriendlythanthegreenvariant,andthegreenrmenters.Thedeterminationofwhichofthefourpossibleequilibriaarisesisdependentuponthexedentrycostandthemarginalcostofmisreportingtheincumbent'slevelofenvironmentalharm.Case1:Thebrownvariantisreportedasbeinglessenvironmentallyfriendlythanthegreenvariant,andthegreenrmenters:Ifthegreenentrantndsitprotabletoenter(i.e.,itcancoveritsxedcostFwiththeprotsearnedinthenalstage),thetwormssharethemarket.FollowingfromEquation 1 ,consumerspurchaseoneunitofthevariantthatyieldsthehighestutility,ortheypurchaseneither.Consumersthathaveastrongdistasteforenvironmentalharmchoosetopurchasefromthegreenentrantatahigherprice,whileconsumerswithlessdistasteforenvironmentalharmchoosetopurchasefromthebrownincumbentatalowerprice.Itisassumedthatthemarketisnotfullycovered;thatis,someconsumers(withaparticularlystrongdistasteforenvironmentalharm)choosenottopurchaseeithervariantofthepollutingproduct.Whenthegreenrmchoosestoenter,andconsumerscorrectlyperceiveitsvariantasthegreenerone,therespectivedemandfunctionsfacedbythegreenandbrownrmaregivenas: qG(pB,pG,^eB,eG)=V)]TJ /F7 11.955 Tf 11.96 0 Td[(pG (eG))]TJ /F7 11.955 Tf 17.43 8.09 Td[(pG)]TJ /F7 11.955 Tf 11.96 0 Td[(pB (^eB)]TJ /F7 11.955 Tf 11.96 0 Td[(eG),and(1) qB(pB,pG,^eB,eG)=pG)]TJ /F7 11.955 Tf 11.96 0 Td[(pB (^eB)]TJ /F7 11.955 Tf 11.96 0 Td[(eG),(1)wherepBandpGarethepricesofthebrownandgreenvariantsrespectively.Thevariable^eBrepresentsthelevelofenvironmentalharmthattheincumbentreportstoconsumers,andeGrepresentstheexogenouslyassignedamountofenvironmentalharmcausedbytheentrant'svariant.Thatis,^eBistheincumbentrm`sfalsereportofenvironmentalharm,whichisassumedtoalwaysbelessthaneB.Thesedemandfunctionscharacterizetheconsumers'responsestothenalstageequilibriumprices. 21

PAGE 22

InStage3,thetwormssimultaneouslysetpricesfortheirrespectivevariants.Withoutlossofgenerality,itisassumedthatthermsdonotincurmarginalcostsofproduction.Protmaximizationissynonymouswithrevenuemaximization.Specically,ifthexedcostsofentryaresmallenoughtofacilitateentry,theentrantandincumbentsolvethefollowingmaximizationproblems,respectively. maxpGG=V)]TJ /F7 11.955 Tf 11.96 0 Td[(pG (eG))]TJ /F7 11.955 Tf 17.43 8.09 Td[(pG)]TJ /F7 11.955 Tf 11.96 0 Td[(pB (^eB)]TJ /F7 11.955 Tf 11.96 0 Td[(eG)pG,and(1) maxpBB=pG)]TJ /F7 11.955 Tf 11.96 0 Td[(pB (^eB)]TJ /F7 11.955 Tf 11.96 0 Td[(eG)pB.(1)TheFirstOrderConditionsforprotmaximizationinCase1showthatthebrownincumbentsellstoconsumerswithlessdistasteforenvironmentalharm.Specically,itsellstoconsumerscharacterizedby2[0,V 4^eB)]TJ /F9 7.97 Tf 6.59 0 Td[(eG]andreceivesapriceofpB=V(^eB)]TJ /F9 7.97 Tf 6.58 0 Td[(eG) 4^eB)]TJ /F9 7.97 Tf 6.58 0 Td[(eG.Thegreenentrantsellstocustomerswithstrongerdistastesforenvironmentalharm,thosecharacterizedby2[V 4^eB)]TJ /F9 7.97 Tf 6.58 0 Td[(eG,V(2^eB+eG) eG(4^eB)]TJ /F9 7.97 Tf 6.59 0 Td[(eG)].ThegreenentrantchargesapriceofpG=2V(^eB)]TJ /F9 7.97 Tf 6.58 0 Td[(eG) 4^eB)]TJ /F9 7.97 Tf 6.58 0 Td[(eG.Thepriceofthegreenvariantistwicethatofthebrownvariant(i.e.,pB=pG 2).Thiscaseariseswhenthexedcostofentryaresufcientlylow.DeterminingwhetherthegreenentrantchoosestoenterinStage2requiresacomparisonofitsexpectedpayoffstoentry(i.e.,post-entryrevenue)anditsxedcostofentry.Itentersprovidedthatthexedcostofentryissufcientlylow.Morespecically,thegreenentrantchoosestoenterif: F4V2(^eB)]TJ /F7 11.955 Tf 11.96 0 Td[(eG)^eB (4^eB)]TJ /F7 11.955 Tf 11.95 0 Td[(eG)2eG.(1)Themaximumxedentrycostthatresultsinentrybythegreenrmisincreasinginthelevelofenvironmentalharmcausedbythebrownincumbent.Itismorelikelythattheentrantcancoveritsxedcostswhenconsumersbelievethattheincumbent'svariantisdirtier.Conversely,ifthebrownrmreducestheperceivedharmofitsproductthroughgreenwash,theentrantndsitmoredifculttoenter.Anygreenwashbytheincumbentservestomakeentrylesslikely. 22

PAGE 23

Interestingly,thegreenrm'srevenuesfalltozerowhentheincumbentreportsitsvariantasbeingequallyasharmfulastheentrant'svariant,implyingthatatsome^eB>eG,thegreenrmisunabletorecoveritsxedcostsofentry.ThisfeatureofthemodelstandsincontrasttoSchmalensee(1982)andHattoriandHigashida(2011a),wherefalseadvertisingcreatesapositiveexternalityforcompetitors.Nosuchexternalityexistsinthismodel.Here,falseadvertisingalwaysreducestheentrant'spost-entryrevenuesthroughheightenedpricecompetition.Case2:Thebrownvariantisreportedasbeinglessenvironmentallyfriendlythanthegreenvariant,andthegreenrmdoesnotenter:Ifthexedcostofentry,F,issufcientlylarge,oriftheentrant'sexpectedpost-entrymarketshareandprotabilityaresufcientlyreduced,thegreenrmisforeclosedfromthemarket.Specically,iftheconditioninEquation 1 doesnothold,thegreenrmdoesnotenter.Thebrownincumbentproducestheonlyavailableproductvariantandservesallconsumerscharacterizedby2[0,V)]TJ /F9 7.97 Tf 6.59 0 Td[(pB ^eB].Theincumbentsetsitsmonopolypricetomaximizeitsprots: maxpBB=V)]TJ /F7 11.955 Tf 11.95 0 Td[(pB ^eBpB.(1)Theincumbentsellsitsvarianttoallconsumerscharacterizedby2[0,V 2^eB]ataprice,pB=V 2.ThebrownincumbentearnsB=V2 4^eB.ComparingtheseprotstothoseearnedbytheincumbentinCase1,itiseasilyseenthatthebrownincumbentexperienceshigherprotswhenitdoesnotfaceacompetitorinthenalstagesofthegame,makingmonopolystatusparticularlydesirable.AsidentiedinEquation 1 ,thisoutcomeisdependentupontheincumbentundertakingasufcientlyhighlevelofgreenwash.Case3:Thebrownvariantisreportedasbeingmoreenvironmentallyfriendlythanthegreenvariant,andthegreenrmdoesnotenter:Itispossiblethattheincumbentchoosestoreportitsvariantaslessharmfulthattheentrant'sgreenervariant(i.e.,^eB
PAGE 24

protablyinStage2.Consequently,when^eBfallsbeloweG,thegreenrmremainsoutofthemarket,similartoCase2.Onlyonce^eBfallssufcientlybeloweGcanthegreenrmenterwiththelowqualityvariant.Whenthebrownrmreportsitsvariantasslightlylessharmfulthanthegreenentrant'svariant,theincumbentrmremainsamonopolist,asitwasinCase2.ItcontinuestochargeapriceofpB=V 2andserveconsumerscharacterizedby2[0,V 2^eB].Asthiscaseoccurswhen^eBtakesalowervaluethaninthepreviouscase,moreconsumersdecidetopurchasethemonopolist'svariant.Case4:Thebrownvariantisreportedasbeingmoreenvironmentallyfriendlythanthegreenvariant,andthegreenrmenters:Asnotedabove,thebrownincumbentcanreportitsvariantasmoreenvironmentallyfriendlythanthegreenrm'svariant.Whenthebrownvariantisperceivedasonlyslightlymorefriendlythanthegreenvariant,thegreenrmisnotabletoenterinStage2,asitspost-entryprotswillnotcoverthexedcostofentry.Oncethebrownvariantisreportedassubstantiallymorefriendlythanthegreenvariant,thegreenrmmaybeabletorecoveritsxedentrycostbyenteringwiththelowqualityvariant.Inthiscase,thebrownvariantisviewedbyconsumersasthehighqualityvariant,butthegreenrmstillenters.Case4differssomewhatfromtheothercases,asitincludesthemarketcongurationinwhichallpotentialconsumerschoosetopurchaseatleastoneofthetwovariants.Inalloftheothercases,therewassomegroupofconsumersthatchosenottopurchaseeithervariantofthepollutingproduct.InCase4,thebrownincumbentmaychoosetoreportalevelofenvironmentalharmsufcientlylowthatitinducesallpotentialconsumerstopurchase.Thisoutcomeisreferredtoasfullcoverage.Itis,however,alsopossiblethatthereportedlevelofharmisnotsufcientlylowtoachievefullcoverage.Here,wemustconsidertwosubcases:Subcase4a:NoFullCoverageandSubcase4b:FullCoverage. 24

PAGE 25

Subcase4a:NoFullCoverage:Unlikethepreviouscases,inthiscase,^eB
PAGE 26

apriceofpG=V(eG)]TJ /F10 7.97 Tf 9.01 0 Td[(^eB) 4eG)]TJ /F10 7.97 Tf 9.01 0 Td[(^eB.Thepriceofthebrownvariantistwicethatofthegreenvariant(i.e.,pG=pB 2).DeterminingwhetherthegreenentrantchoosestoenterinStage2,when^eB
PAGE 27

SimilartotheanalysispresentedforCase3,thetwormssimultaneouslysetpricesfortheirrespectivevariants,andtheincumbent'sproductisincorrectlybelievedtobethehighqualityvariant.Theentrantandincumbentsolvethefollowingmaximizationproblems,respectively: maxpGG=pB)]TJ /F7 11.955 Tf 11.96 0 Td[(pG eG)]TJ /F8 11.955 Tf 15.09 0 Td[(^eBpG.(1) maxpBB=1)]TJ /F7 11.955 Tf 13.15 8.09 Td[(pB)]TJ /F7 11.955 Tf 11.95 0 Td[(pG eG)]TJ /F8 11.955 Tf 15.09 0 Td[(^eBpB,and(1)TheFirstOrderConditionsforprotmaximizationinCase3showthatthebrownincumbentsellstoconsumerswithmoredistasteforenvironmentalharm.Specically,itsellstoconsumerscharacterizedby2[1 3,1]andreceivesapriceofpB=2(eG)]TJ /F10 7.97 Tf 9.02 0 Td[(^eB) 3.Thegreenentrantsellstocustomerswithweakerdistastesforenvironmentalharm,thosecharacterizedby2[0,1 3].ThegreenentrantchargesapriceofpG=(eG)]TJ /F10 7.97 Tf 9.01 0 Td[(^eB) 3.Theincumbentselectsalevelofgreenwashforitsproduct,andtheintensityofthisfalseadvertisingdeterminestheequilibriummarketarrangement.Fourpossiblecaseswereexploredabove:Case1)Thebrownvariantisreportedasbeinglessenvironmentallyfriendlythanthegreenvariant,andthegreenrmenters;Case2)Thebrownvariantisreportedasbeinglessenvironmentallyfriendlythanthegreenvariant,andthegreenrmdoesnotenter;Case3)Thebrownvariantisreportedasbeingmoreenvironmentallyfriendlythanthegreenvariant,andthegreenrmdoesnotenter;andCase4)Thebrownvariantisreportedasbeingmoreenvironmentallyfriendlythanthegreenvariant,thegreenrmdoesenters.Case4isseparatedintotwosubcases,oneinwhichfullcoverageisnotachievedandoneinwhichitis.Dependinguponthegreenrm'scostofentryandtheincumbent'smarginalcostofmisrepresentingitsproduct,oneofthesefourmarketstructuresarisesinthegame'sSubgamePerfectNashEquilibrium.Ifthexedcostofentryislow,andgreenwashingislow,Case1arises,andtheincumbentchargesalowerpricethantheentrantandsellstoconsumerswithless(dis)tasteforenvironmentalharm.Asthexedcostsofentryincreases,and/orthe 27

PAGE 28

amountofgreenwashingincreases,Case2orCase3arisesinequilibrium,andtheincumbentrmservesasamonopolist.Ifthexedcostsofentryaresufcientlylow,andgreenwashingissufcientlypervasive,Case4arisesinequilibrium,andthegreenrmwillenterasthelowqualityrm.TheentrychoiceisdeterminedbyConditions 1 and 1 ,whiletheamountofoptimalgreenwashisdeterminedbythemarginalcostofgreenwash. 1.3.2Stage1:GreenwashingInStage1,thebrownincumbentchoosesthelevelofgreenwashtoapplytoitsvariant.Itreportstoconsumersthatitsproductislessharmfulthanitactuallyis,deceitfullyincreasingthedemandforitsvariant.9Inthismodel,consumersbelievethattheincumbent'svariantiscleanerthanitactuallyisandmaketheirpurchasedecisionsaccordingtothatfalsebelief.Themodelassumesthatoncegreenwashhasbeenundertakenbytheincumbentrm,neithertheentrantnorconsumersareabletoidentifytheactualamountofharmcausedbytheincumbent.Thisassumptionisbasedonthenotionthatsuchinformationwouldoftenbeunrecoverableinreality.Simplylookingatorconsumingaproductdoesnotrevealtheamountofharmcausedbyitsproduction.Whenaconsumerpurchasesanorganicappleinthesupermarket,forexample,theyarenotabletoverifythattheapplewas,infact,organicallyproduced.Rather,theyrelyontheprovisionofinformationbytheproducerandaremadetobelievetheproducer'sclaim.Manypollutingproductssharethischaracteristic;theyare,inessence,credencegoods.AswasdoneinSubsection3.1,fouruniquemarketarrangements,acrosswhichthemarketstructuresubstantivelydiffers,thatcouldariseinthenalstageofthe 9Iassumethattheexogenouslyassignedlevelsofactualenvironmentalharmaresufcientlydifferentiatedthattheincumbentdoeswishtoundertakesomelevelofgreenwash.Iprovidethenecessaryparameterrestrictionbelow. 28

PAGE 29

gamemustbeconsidered.Thedeterminationofwhichofthesefourmarketstructuresarisesisdependentuponthexedentrycostandthemarginalcostofmisreportingtheincumbent'slevelofenvironmentalharm.ThissubsectioncharacterizestheSubgamePerfectNashEquilibriumcandidatesineachmarketstructure,byfocusingontheincumbent'srststagechoiceofgreenwashing.Particularattentionispaidtohowtheincumbentrespondstothepresenceofdiscontinuitiesinitsprotfunctionthatarisefromalteringthemarketstructure.Case1:Thebrownvariantisreportedasbeinglessenvironmentallyfriendlythanthegreenvariant,andthegreenrmenters:Ifthebrownincumbentcorrectlyreportsitsvariantasmoreharmfulthantheentrant'sgreenervariant,andiftheentrantndsitprotabletoenter(i.e.,itsxedcostFissufcientlylowtofacilitateenetry),thetwormssharethemarket.Moreover,theincumbent'svariantiscorrectlyviewasthelowqualityvariantbyconsumers.SolvingtheFirstOrderConditionsforpricingobtainedinEquation 1 andEquation 1 ,theStage1protmaximizationprobleminCase1isdescribedby: max^eBB=V2(^eB)]TJ /F7 11.955 Tf 11.95 0 Td[(eG) (4^eB)]TJ /F7 11.955 Tf 11.96 0 Td[(eG)2)]TJ /F7 11.955 Tf 13.15 8.08 Td[(k(eB)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB)2 2.(1)Theprotmaximizingreportwhentheentrantentersandtheincumbentdoesnotreportitsvariantaslessharmfulthantheentrant'svariant,^eB,mustsatisfy: V2(4^eB)]TJ /F8 11.955 Tf 11.95 0 Td[(7eG) (4^eB)]TJ /F7 11.955 Tf 11.95 0 Td[(eG)3=k(eB)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB).(1)Here,theanalysisisrestrictedtocasesinwhicheB7eG 4,asviolatingsucharestrictionwouldimplythatthebrownrmdoesnotwishtogreenwashitsproduct.10Itisapparentthatwithinthiscase,reducingthemarginalcostofmisleadingconsumers 10Violatingthisrestrictionwouldimplythatthebrownincumbentcouldprotablyreportitsvariantasmoreharmfulthanitactuallyis.Thisfeaturearises,becauseproductsthatarelessdifferentiatedresultinmoreintensepricecompetitioninthenalstage. 29

PAGE 30

resultsinmoregreenwash.Withoutanyshiftinmarketstructure,reducingthecostofgreenwashresultsinmoregreenwash.Additionally,anyreductioninthecostofgreenwashresultsinmoreenvironmentalharm.Thelevelofenvironmentalharmcausedbyeachproductisobtainedfromtheproductofthatproduct's(actual)perunitharmandthequantityproducedofthatproduct.DeningtheaggregatelevelofenvironmentalharmasPQiei,overallactiverms,thelevelofaggregateenvironmentalharmthatarisesinCase1isgivenas: H=V(2^eB+eB) 4^eB)]TJ /F7 11.955 Tf 11.95 0 Td[(eG.(1)Itiseasytoseethattheaggregatelevelofenvironmentalharmisdecreasingas^eBincreases,orenvironmentalharmincreasesastheincumbentgreenwashesitsproduct.Untiltheincumbentgreenwashesitsproductenoughtoforecloseentry,aggregateenvironmentalharmisunambiguouslyincreasingasthereportedlevelofharmfalls.Moreover,therateatwhichaggregateharmincreasesintheamountofgreenwashisincreasing.Consumerwelfareprovidesanadditionaldimensiononwhichtoconsidertheeffectoffalseadvertising.Withinthiscase,greenwashingservestointensifypricecompetition,whichbenetsconsumers.DeningconsumersurplusasCSR^0V)]TJ /F6 11.955 Tf 12.27 0 Td[(eB)]TJ /F7 11.955 Tf 12.27 0 Td[(pBd+R~^V)]TJ /F6 11.955 Tf 12.22 0 Td[(eG)]TJ /F7 11.955 Tf 12.23 0 Td[(pGd,whereconsumerscharacterizedby2[0,^]purchasethevariantofferedbythebrownincumbentandconsumerscharacterizedby2[^,~]purchasethevariantofferedbythegreenentrant,analyzingtheeffectofgreenwashonconsumerwelfareisstraightforward.Reductionsintheincumbent'sreportedlevelofenvironmentalharmbelowitsactuallevelofharminitiallyincreaseconsumersurplus.Consumer'sbenetfromfalseadvertising,whenever^eB>4eB)]TJ /F10 7.97 Tf 6.59 0 Td[(3eG 16;however,oncethereportedlevelofharmfallsbelowthisthreshold,greenwashbeginstoharmconsumers.ItisnotclearwhethergreenwashiseverabletoharmconsumersurpluswithinCase1,becausehighxedcostsofentrymayimplythattheconditionaboveisonlyviolatedwhenthe 30

PAGE 31

incumbent'slevelofgreenwashforeclosesthemarket.Asaconsequence,Consumersurplusproducestwocandidatemaximaforconsumersurplus. Corollary1. When^eB>eG,andtheentrantisabletoprotablyenter:(i)Anyincreaseinthemarginalcostofgreenwashreducestheamountofgreen-washundertakenbytheincumbentatadecreasingrate.(ii)Anyincreaseinthereportedlevelofenvironmentalharmbytheincumbentdecreasesaggregateenvironmentalharmatadecreasingrate.(iii)thereportedlevelofharmbytheincumbentthatmaximizesconsumersurplusislessthaneB.When^eB=4eB)]TJ /F10 7.97 Tf 6.58 0 Td[(3eG 16resultsinentrybythegreenrm,^eB=4eB)]TJ /F10 7.97 Tf 6.59 0 Td[(3eG 16maximizesconsumersurplus.When^eB=4eB)]TJ /F10 7.97 Tf 6.58 0 Td[(3eG 16doesnotresultinentrybythegreenrm,areportedlevelofharmjustabovetheentrythresholdlevelofharmmaximizesconsumersurplus.Reducingthemarginalcostofmisleadingconsumersalsomakesitmorelikelythattheincumbentwillforeclosetheentrantfromthemarket,astheentrant'spotentialrevenuesfallastheincumbentgreenwashesitsvariant.Dening~F(^eB)4V2(^eB)]TJ /F9 7.97 Tf 6.59 0 Td[(eG)^eB (4^eB)]TJ /F9 7.97 Tf 6.58 0 Td[(eG)2eGasthethresholdlevelofentrycost,anyxedentrycostF>~F(^eB)isprohibitivelyhighfortheentranttorecoverpost-entry.Greenwashingbythebrownincumbentreducesthelikelihoodofentrybythegreenentrant,becauseitreducestheentrycostthreshold.~F(^eB)isincreasingin^eB,ordecreasingintheamountofgreenwash,eB)]TJ /F8 11.955 Tf 15.93 0 Td[(^eB.Thatis,whenthebrownincumbentreportsitsproductaslessharmfultoconsumers,thethresholdlevelofxedcostsforentryforeclosurefalls,anditbecomesmorelikelythatthexedentrycostisprohibitivelyhigh.Asthemarginalcostofgreenwashingdecreases,theprot-maximizinglevelofgreenwashincreases.Theseheightenedgreenwasheffortsbytheincumbentmakeentrylesslikely.Oncetheamountofgreenwashissufcienttoforeclosethemarketfromentrybythegreenrm,theincumbent'sprotsexhibitanupwardjump.Thatis,notsurprisingly,theincumbent'sprotsexperienceanupwardjumpwhenitsgreenwash 31

PAGE 32

effortssecuremonopolystatus.Additionally,asthemarginalbenetofgreenwashalsojumpsupwhenentryisprevented,thereisanupwardjumpintheprot-maximizingamountofgreenwashwhenthemarketarrangementmovesfromthatdescribedbyCase1tothatdescribedbyCase2.Case2:Thebrownvariantisreportedasbeinglessenvironmentallyfriendlythanthegreenvariant,andthegreenrmdoesnotenter:Ifthexedcostofentry,F,issufcientlylarge,thegreenentrantisforeclosedfromthemarket.Specically,iftheconditioninEquation 1 doesnothold,thegreenrmcannotprotablyenter.Thebrownincumbentproducestheonlyproductvariantandservesallconsumerscharacterizedby2[0,V 2^eB].Theincumbentsetsitsmonopolypricetomaximizeitsprots: max^eBB=V2 4^eB)]TJ /F7 11.955 Tf 13.15 8.09 Td[(k(eB)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB)2 2.(1)Theprotmaximizingreportwhentheentrantdoesnotenter,andtheincumbentdoesnotreportitsvariantasmoreharmfulthantheentrant'svariant,^eB,isgivenbythefollowingFirstOrderCondition: V2 4^eB2=k(eB)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB).(1)Itiseasilyshownthatincreasesinthemarginalcostofgreenwashingincreasestheincumbentrm'sprot-maximizingreportsolongasthosereductionsactuallyleadtomoregreenwashing.Inordertoruleoutrangesofkoverwhentheprotmaximizingreportisheldconstant,wemustcarefullyconsidertheincumbent'sgreenwashingbehaviorintheneighborhoodofthereportthatdetersentry(betweenCase1andCase2).Theincumbent'sprotsjumpupwardwhenitsreportdetersentry.Moreimportantly,themarginalbenetofgreenwashalsojumpsup.Whentheincumbent'sreportissufcientlylowtodeterentry,itsprotmaximizingreportjumpsdown.Asadownwardjumpinthereportedlevelofharmdoesnotviolatetheentrydeterrencecondition,the 32

PAGE 33

protmaximizinglevelofreportedharmisnotstrategicallyheldconstantoveranyrangeofkintheneighborhoodoftheentrythreshold. Corollary2. If^eB>eG,whengreenwashissufcientlystrongtoforeclosethemarkettoagreenerentrant,theincumbent'sprotsexhibitanupwardjump,andthemarginalbenetofgreenwashexhibitsanupwardjump.Together,thesefeaturesimplythattheprotmaximizingreportjumpsdown.Theamountofaggregateenvironmentalharmcausedbythemonopolistincumbent'sproductvariantisgivenasH=VeB 2^eB.Increasesintheincumbent'slevelofreportedenvironmentalharmreduceaggregateenvironmentalharmasadecreasingrate.Thisimpliesthatgreenwashbyanincumbentincreasesaggregateharmatanincreasingrate.Additionally,anyincreaseinthemarginalcostofgreenwashreducestheaggregatelevelofenvironmentalharmatadecreasingrate.Case3:Thebrownvariantisreportedasbeingmoreenvironmentallyfriendlythanthegreenvariant,andthegreenrmdoesnotenter:Thereisnorestrictionregardingtheincumbent'sabilitytoreportthatitsvariantislessharmfulthanthepotentialentrant'svariant.Theincumbentmayreportthat^eBislessthaneG.Iftheincumbentreportsitsvariantasonlyslightlylessharmfulthanthepotentialentrant'svariant,themarketremainsforeclosed,effectivelymaintainingthemonopolystructureoutlinedaboveforCase2.Onlywhentheincumbent'svariantisreportedassubstantiallycleanerthantheentrant'svariantcantheentrantrecoverthexedcostsofentry.Thisoccurs,becauseincreasedproductdifferentiationsoftenspost-entrypricecompetition,allowingthenentranttorecoveritsxedentrycostbysellingthelowqualityvariant.Astheincumbent'sgreenwasheffortsbecomesufcientlystrongtoalterthemarketstructureandfacilitateentrybythegreenerentrant,theincumbent'sprotsexhibitadownwardjump,butthemarginalbenetofgreenwashjumpsup.Again,asinCase2,theincumbentdoesnothaveanincentivetostrategicallyavoidCase4bymaintainingaconstantlevelofgreenwashthatisjustabovethelevelthatwouldfacilitateentry. 33

PAGE 34

Whentheincumbent'slevelofreportedharmissetsufcientlylowtoforeclosethemarket(betweenCase1andCase2),theprot-maximizingreportjumpsdownward.Similarly,whentheincumbent'slevelofreportedharmissetsufcientlylowtofacilitateentryintothemarket(betweenCase3andCase4),theprot-maximizingreportalsojumpsdownward.Thereisnorangeofkwithinthemonopolymarketstructureoverwhichtheprotmaximizinglevelofgreenwashisheldconstant,becausethemarginalbenetofgreenwashisconsistentwiththeentrythresholds.Withinthemonopolymarketstructure,anyincreaseinkincreasestheincumbent'sprot-maximizingreportatadecreasingrate.Whenevertheincumbentrmisabletosecureamonopoly,increasingitsmarginalcostofgreenwashreducesitsprot-maximizinglevelofgreenwashatadecreasingrate.Thishasstraightforwardimplicationsforaggregateenvironmentalharmandconsumersurplus.Overtherangeofkthatsupportamonopolymarketstructure,aggregateenvironmentalharmisgivenasH=VeB 2^eB.Regardlessofwhether^eB>eG,aggregateenvironmentalharmfallsatadecreasingrateaskincreases.Consumersurplusjumpsdownwheneverthemarketisforeclosed.Additionally,withinthemonopolymarketstructure,anyincreaseinthemarginalcostofgreenwashresultsinincreasesconsumerwelfare.Here,consumersurplusisdenedasCSR^0V)]TJ /F6 11.955 Tf 12.57 0 Td[(eB)]TJ /F7 11.955 Tf 12.57 0 Td[(pBd,whereconsumerscharacterizedby2[0,^]purchasethevariantofferedbythebrownincumbent.Asthereportedlevelofharmincreases,withinthismonopolymarketstructure,consumersaremadebetteroff,unambiguously.Higherlevelsofreportedharmreducepricesandinefcientproductswitching. Corollary3. Whengreenwashbytheincumbentpreventsprotableentrybythegreenrm:(i)Anyincreaseinthemarginalcostofgreenwashreducestheamountofgreen-washundertakenbytheincumbentatadecreasingrate. 34

PAGE 35

(ii)Anyincreaseinthemarginalcostofgreenwashreduceaggregateenvironmentalharmatadecreasingrate.(iii)Consumersurplusjumpsdownwheneverentryisforeclosed,regardlessofwhether^eB>eG.Additionally,withintherangeofmarginalcostsofgreenwashthatgivesrisetoamonopolymarketstructure,increasesinthemarginalcostofgreenwashincreaseconsumersurplusatadecreasingrate.Elevatedmonopolyprotsbytheincumbentgiverisetodiscontinuitiesintheincumbent'sprotfunction.Theprot-maximizingreportsintheneighborhoodsoftheentrythresholds,however,donotconictwiththeentrycostconditions.Asaconsequence,theFirstOrderConditiongiveninEquation 1 fullycharacterizesgreenwashingbehaviorwithinthemonopolymarketstructure.Additionally,withinthismarketstructure,anyincreaseinthemarginalcostofgreenwashingreducesaggregateenvironmentalharmatadecreasingrateandincreasesconsumersurplusatadecreasingrate.Case4:Thebrownvariantisreportedasbeingmoreenvironmentallyfriendlythanthegreenvariant,andthegreenrmenters:Itispossiblethattheincumbentchoosestoreportitsvariantaslessharmfulthantheentrant'sgreenervariant(i.e.,^eB
PAGE 36

levelofharmissufcientlyhightoavoidfullmarketcoverage.Thatis,someconsumerschoosenottopurchaseeithervariant,aseventheincumbent'svariantisperceivedastooharmfulbysomeconsumers.Inthesecondsubcase,theincumbent'sreportedlevelofharmissufcientlylowtodrawallconsumersintothemarket,andfullcoverageisachieved.Subcase4a:NoFullCoverage:Whenthebrownincumbentjumpsitsrival(intermsofreportedenvironmentalharm),itisperceivedasofferingthehighqualityproduct.Ifthejumpissufcientlylarge,thegreenrmenters,resultinginaduopolymarketstructure.Ifthatjumpisnotsufcientlylargetodrawallconsumersintothemarket(i.e.,toachievefullmarketcoverage),theincumbent'sStage1maximizationproblemisgivenasfollows: max^eBB=4V2(eG)]TJ /F8 11.955 Tf 15.08 0 Td[(^eB)eG (4eG)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB)2^eB)]TJ /F7 11.955 Tf 13.15 8.09 Td[(k(eB)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB)2 2.(1)Whenthereportedlevelofharmissufcientlylowtofacilitateentry,theincumbent'sprotsjumpdown,butthemarginalbenetofgreenwashjumpsup.Whenentry(inCase4)arises,theprotmaximizingreportjumpsdown.Asintheothercases,increasesinthemarginalcostofgreenwashingincreasetheprot-maximizingreportofenvironmentalharmbytheincumbentatadecreasingrate.Thatis,@^eB @k>0and@2^eB @k2<0.Whentheincumbentservesthemarketwiththehighqualityvariant,andtheentrantservesthemarketwiththelowqualityvariant,amarketthatisnotfullycoveredresultsinanamountofaggregateenvironmentalharm: H=VeG(2eB+^eB) (4eG)]TJ /F8 11.955 Tf 15.08 0 Td[(^eB)^eB.(1)Itiseasytoshowthattheaggregatelevelofenvironmentalharmisdecreasingas^eBincreases,orenvironmentalharmincreasesastheincumbentgreenwashesitsproduct.Additionally,as@^eB @k>0,increasingthemarginalcostofgreenwashingleadstoreductions(atadecreasingrate)intheaggregatelevelofenvironmentalharm.Similar 36

PAGE 37

toCase1,withinamarketthatisarrangedasperCase4andisnotfullycovered,smallincreasesinthemarginalcostofgreenwashalwaysincreasetheincumbent'sreportatadecreasingrateandreduceharmatadecreasingrate.Whentheincumbentmisreportsitsvariantaslessharmfulthantheentrant'svariant,greenwashalwaysharmsconsumers.UnlikeinCase1wheregreenwashbenetedconsumersgreenwashinCase4servestoincreaseperceivedproductdifferentiationandsoftenpricecompetition.Consumeralsomakeinefcientpurchasedecisions,withconsumerswiththemostsensitivitytoenvironmentalharmpurchasingthemoreharmfulvariant. Corollary4. When^eB
PAGE 38

alreadypurchasingoneofthetwoavailablevariants.Inthiscase,theincumbentsellstothe2 3ofthecustomerswiththehigherdistastesforenvironmentalharmandmaximizesprotsaccordingto: max^eBB=4(eG)]TJ /F8 11.955 Tf 15.08 0 Td[(^eB) 9)]TJ /F7 11.955 Tf 13.16 8.08 Td[(k(eB)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB)2 2.(1)Theincumbentrm'sprotsjumpupward,butthechangemarginalbenetofgreenwashingisambiguouswhenfullcoverageisachieved,relativetoSubcase4a.11Fullcoveragereducestheneedfortheincumbentrmtoconsidermarginalconsumers(forthemarket).Thisresultsinreduceddownwardpricepressure,post-coverage,whichleadsprotstojumpupwardwhenfullcoverageisachieved.Theeffectoffullcoverageontheincumbent'sprot-maximizinglevelofreportedharmislessclear,becauseitisnotclearwhetherthemarginalbenetofgreenwashisgreaterpre-coverageorpost-coverage.Theincumbent'sprotsunambiguouslyjumpupwardwhenthemarketbecomesfullycovered,buttheeffectofgreenwashislessclear.Deningthepre-coveragemarginalbenetofgreenwashasMBNCandthepost-coveragemarginalbenetofgreenwashasMBFC,wemustconsiderboththepossibilitythatMBNC>MBFCandMBNCMBFC,forarangeofvaluesofkclosetothecoveragethreshold,theprot-maximizinglevelofreportedharmisheldconstantjustbelowthecoveragethresholdlevel.IfMBNC
PAGE 39

foreachofthetwopossibilities.Regardlessofwhetherreportedharmisincreasingorheldconstant,theaggregatelevelofenvironmentalharmisconstantwhenthemarketisfullycovered.Aggregateenvironmentalharmisgivenas: H=2eB+eG 3(1)Aggregateharmremainsconstantacrossallvaluesofkthatgiverisetofullmarketcoverage.Whenthemarketisfullycovered,thetotalmarketoutputandthemarketsharesofthetwormsremainconstant.Consequentially,aggregateharmcanneitherbeincreasednordecreased.Whentheincumbentmisreportsitsvariantaslessharmfulthantheentrant'svariant,andthemarketisfullycovered,greenwashdecreasesconsumersurplusataconstantrate.Whenthemarketisfullycovered,themarketsharesofeachrmareconstant;changesinthereportedlevelofharmonlyaffectthepriceschargedbyeachrm.Whentheincumbentrmreducesitsreportedlevelofharm,thevariantsareperceivedasmoredifferentiated,andpricesrise.Asaconsequence,consumersarealwaysharmsbygreenwashwhenthemarketisfullycovered. Corollary5. When^eB
PAGE 40

respondstothosestructuralchanges.Increasesinthemarginalcostofgreenwashingarecapturedbyincreasesintheparameterk.Thisparametermayrepresenttheactualcostofagreenwashingcampaign,oritmaybeinterpretedascapturingthegullibilityofconsumers.Itislesscostlytodeceivegullibleconsumers.Theguredemonstratesthattheamountofgreenwashisgenerallydecreasingatadecreasingrateasthemarginalcostofgreenwashingrises,withsomeapparentexceptions.Therearetwopossiblerangesofkoverwhichtheoptimallevelofgreenwashisheldconstantbotharisingwhenthemarketisfullycovered.Therearethreepossiblediscontinuousjumpsatwhichincreasesinthemarginalcostofgreenwashresultinlessgreenwashing,buttherearenoupwardjumpsingreenwashefforts.Thatis,increasesinthemarginalcostofgreenwashmaynotalwaysreducegreenwashing,buttheyneverincreaseit.Moreover,whenthemarketisnotfullycovered,anyincreaseinthemarginalcostofgreenwashreducesgreenwasheffortsbytheincumbent. Lemma1. Increasesinthemarginalcostofgreenwashmaynotalwaysreducegreen-washing,buttheyneverincreaseit.Ifthemarketisnotfullycovered,increasingthemarginalcostofgreenwashalwaysreducesgreenwashingeffortsbytheincumbent.Forverylowvaluesofk,themarketisfullycovered,meaningthatallconsumerschoosetopurchaseoneunitofoneofthetwoavailablevariants.Fullcoverageisobtainedwhentheincumbentsreportedlevelofenvironmentalharmissufcientlylowtodrawallconsumersintothemarket.13Thethresholdreport(forfullcoverage)bytheincumbentispositiveandarisesatapositivemarginalcostofgreenwash.Withinthisrangeofverylowmarginalcostsofgreenwashthatgiverisetoafullycovered jumpsup,theprot-maximizingreportwouldjumpdownandnotbeheldconstantforarangeofkintheneighborhoodofthecoveragethreshold.13Whenthemarketisfullycovered,theentrantisabletoenter,butconsumersincorrectlyviewtheincumbent'sproductasthehighqualityvariant. 40

PAGE 41

Figure1-1. TheEffectofkonReportedHarm market,threepossiblerelevantsubrangescanbeidentied.Forextremelylowvaluesofk,theincumbent'sreportisheldconstantatzero.Withinthissubrange,theprotmaximizingreportisnegative.Inthismodel,theincumbentisnotallowedtoreportitsproductasenvironmentallybenecial.Thus,forthoseextremelylowvaluesofk,thereportedlevelofharmisequaltozero,implyingthat^eB=eB.Oncekrisesabovethepointatwhichtheincumbentchoosestoreportapositivelevelofenvironmentalharm,marginalincreasesinkreducegreenwashingatadecreasingrate.Eventually,inanefforttomaintainafullycoveredmarket,theincumbentmayholditsgreenwashingconstantatthelevelthatjustcoversthemarket.Thisnalsubrangeonlyarisesisthe 41

PAGE 42

marginalbenetofgreenwashjumpsupwhenthemarketisnolongerfullycovered;otherwise,thisnalsubrangedoesnotarise.Asaconsequence,withinthefullycoveredmarketarrangement,thereexistthreepossiblesubrangesofk(listedinascendingorder):asubrangeoverwhichthereportedlevelofenvironmentalharmisheldconstantatzero,asubrangeoverwhichmarginalincreasesinkincreasethereportedlevelofenvironmentalharmatadecreasingrate,andasubrangeoverwhichthereportedlevelofenvironmentalharmisheldconstantatapositivelevel.Therearenodiscontinuousjumpsinthereportedlevelofenvironmentalharmwithinthefullycoveredmarketstructure.Aswithlowervaluesofk,whenthemarketisfullycovered,thereisarangeofmarginalcostsofgreenwashingthatgiverisetoaduopolymarketinwhichconsumersincorrectlyperceivetheincumbent'sproductasthehighqualityvariant,butthemarketisnotfullycovered.Thatis,thereportedlevelofenvironmentalharmissufcientlyhightoleadthemostenvironmentallyconsciousconsumerstoabstainfrompurchasingeithervariant.Withintherangeofmarginalcostsofgreenwashthatgivesrisetosuchamarketstructure,anyincreaseinthemarginalcostofgreenwashreducesgreenwashingbytheincumbentatadecreasingrate.Formarginalcostsofgreenwashthatresultintheincumbentchoosingareportedlevelofenvironmentalharmclosetotheentrant'sactuallevelofharm,theentrantcannotrecoverthexedcostsofentry,andtheincumbentservesthemarketasamonopolist.Thismonopolymarketstructurecanarisewhen^eB>eGor^eB
PAGE 43

Forhighermarginalcostsofgreenwashing,greenwashbytheincumbentistooweaktodeterentrybythegreenerrm,andaduopolymarketinwhichconsumerscorrectlyviewtheincumbent'sproductasthelowqualityvariantarises.Withinthismarketstructure,anyincreaseinthemarginalcostofgreenwashreducesgreenwashingbytheincumbentatadecreasingrate.Abrownincumbentmonopolistcangreenwashitsproducttoincreaseitsownmarketshareintheverticallydifferentiatedmarket.Byfalselyportrayingitsproductaslessenvironmentallyharmfulthanitactuallyis,theincumbentattractsconsumerswithhigherdistastesforenvironmentalharmthanwouldnormallybuyitsproductvariant.Giventheverticalnatureofgreenproductsandpositiveentrycosts,greenwashbyanincumbenthastheabilitytoalterthemarketstructure.Thissectiondemonstratesthatrelativelysmallamountsofgreenwashmayforeclosethemarketanddeteragreenerrmfromentering.Evengreateramountsofgreenwashmayfacilitateentrybythegreenrm,butwithitsvariantperceivedaslowquality.Theprot-maximizinglevelofgreenwashisdrivenbythecostofgreenwashing,andthissectionoutlineshowoptimalgreenwashrespondstochangesinthemarginalcostofgreenwashing.AsnotedbyLemma 1 ,increasingthemarginalcostofgreenwashneverresultsinmoregreenwash,andifthemarketisnotfullycovered,italwaysreducesit.Greenwashingdoesnotsimplyaffectmarketstructure;ithasimportantimplicationsforconsumerwelfareandtheamountofaggregateenvironmentalharmthatarises. 1.4ImplicationsforConsumerWelfareandEnvironmentalHarmGreenwashingbyanincumbentmonopolistservestomakeitsproductmoreappealingtocustomers.Anincumbentrmgreenwashesitsproducttoincreaseitspotentialmarketshareandincreaseitsprots.Ifgreenwashingissufcientlystrong,itcanforeclosethemarkettoanentrantthatmayofferagreenerproductorevenallowentry,butmisrepresentitsdirtyproductvariantasgreenerthantheentrant'svariant.Theabilityofgreenwashtoalterthemarketstructurehasimportantimplicationsfor 43

PAGE 44

consumerwelfareandenvironmentalharm.Someoftheseeffectsarestraightforward,whereassomearelessobvious.Understandingtheseeffectsisparticularlyusefulformarketregulatorsandenvironmentalactivistswhoseektoreducegreenwashingbehaviorbyalteringitscosts.Consumerwelfare,acommoninterestformarketregulators,isgivenasthe(indirect)utilitythateachconsumerenjoysbymakingtheirequilibriumconsumptionchoice,summedacrosstheconsumersthatchoosetopurchaseoneoftheavailablevariants.EachconsumerreceivestheindirectutilitydescribedbyEquation 1 forpurchasingaparticularvarianti;thelevelofutilityisdeterminedbythepriceandactualenvironmentalharmcausedbythatvariant.The(falsely)reportedlevelofharmonlyaffectsaconsumer'swelfarethroughthereport'seffectonpurchasedecisions.Figure1-2illustratestheeffectofgreenwashingontheoveralllevelofconsumersurplus.Itplotsconsumersurplusagainstthemarginalcostofgreenwashing.ConsistentwiththeresultsinHattoriandHigashida(2011b),whengreenwashdoesnotalterthemarketstructure(andentryoccurs),greenwashmaybenetconsumers.Thisresultarises,becausegreenwashingservestointensifypricecompetitionbetweenthetworms.Reducingitsreportedlevelofenvironmentalharminthedirectionoftheentrant'sactuallevelofharmdecreasestheamountofdifferentiationbetweenthevariants,anddiminisheddifferentiationresultsinheightedpricecompetitioninthenalstageofthegame.ItispossiblethatConsumerSurplusachieveslocalmaximumwithintheCase1marketstructureifthetwoproductsaresufcientlydifferentiated.Oncegreenwashissufcientlystrongtoalterthemarketstructureeitherbysecuringmonopolystatusfortheincumbentorgivingrisetoaduopolyinwhichtheincumbentoffersthehighqualityproductgreenwashgenerallyservestoharmconsumersatthemargin.Intheseinstances(e.g.,inCases2-4),greenwashonlyservestomisleadconsumersintomakinglessefcientconsumptionchoices,andintensiedpricecompetitionisnotabletocompensateforthatnegativewelfareeffect. 44

PAGE 45

Figure1-2. TheEffectofkonConsumerSurplus Additionally,whentheduopolyinwhichtheincumbentoffersthehighqualityproductarises,greenwashreducespricecompetitionbyheighteningthelevelofdifferentiationbetweentheavailablevariants.Ifgreenwashaltersthemarketstructure,itgenerallyreducesconsumerwelfare.Theonlyexceptionariseswhengreenwashissufcientlystrongtofacilitateentrybytheentrantasthelowqualityrm.Whengreenwashissufcientlylargetofacilitateentrybythegreenrm(butashavingthelowqualityvariant),consumersurplusjumpsupward.Thisdiscontinuousupward(whenviewedinthecontextofgreenwash,asopposedtoharm)jumpinwelfarearisesasaconsequenceofincreasedpricecompetition.Consumersbenetfromentry 45

PAGE 46

bythepotentialentrant,evenwhenitsproductisincorrectlyviewedasthelowqualityvariant.Entryintroducescompetitioninthepricingstageofthegame.Theglobalmaximumforconsumersurplusariseswithsomepositiveamountofgreenwash,asgreenwashinitiallybenetsconsumers.Relativelysmallamountsofgreenwashalwaysbenetconsumers,asgreenwash(withinCase1)softenspricecompetition.Thisfeatureofthemodelimpliesthatsomeamountofgreenwashisdesirablefromtheconsumers'perspective.Giventheresponseofconsumerwelfaretogreenwashadvertisingandtherelatedchangesinthemarketstructure,therearethreecandidatesfortheglobalmaximumlevelofconsumersurplus:1)when^eB=4eB)]TJ /F10 7.97 Tf 6.59 0 Td[(3eG 16resultsinentrybythegreenrm,^eB=4eB)]TJ /F10 7.97 Tf 6.59 0 Td[(3eG 16maximizesconsumersurpluswithinCase1,2)when^eB=4eB)]TJ /F10 7.97 Tf 6.59 0 Td[(3eG 16doesnotresultinentrybythegreenrm,areportedlevelofharmjustabovetheentrythresholdlevelofharmmaximizesconsumersurplusinCase1,and3)areportedlevelofharmjustbelowtheentrythresholdlevelofharmmaximizesconsumersurplusinCase4. Lemma2. Theglobalmaximumforconsumersurplusariseswhenapositiveamountofgreenwashisundertakenbytheincumbentrm,andtheamountofgreenwashdoesnotforeclosethemarketfromentry.Thexedcostsofentryalsohaveimportantimplicationsfortheoveralllevelofconsumersurplus.Increasingtheamountofxedcostsdecreasesconsumersurpluswhenitchangesthepotentialentrant'sentrydecision.Asconsumersalwaysbenetfromentrythroughincreasedcompetition(regardlessofhowtheincumbentpositionsitsvariant),increasingthexedcostsofentryisalwaysharmfulforconsumers.Aseparateandoften,equallyormoreimportantissueformarketregulatorsandenvironmentalactivistsiswhethergreenwashresultsinincreasedordecreasedlevelsofaggregateenvironmentalharm.AsinHeijenandSchoonbeek(2008),environmentalharmisdenedasPQiei,overallirms.Figure1-3illustratesthe 46

PAGE 47

effectofgreenwashonaggregateenvironmentalharm.Itplotsaggregateenvironmentalharmagainstthemarginalcostofgreenwashing. Figure1-3. TheEffectofkonAggregateEnvironmentalHarm Increasinggreenwashgenerallyresultsinincreasedaggregateenvironmentalharm,withonepossibleexception.AsshownbyLemma1,increasesinthemarginalcostofgreenwashdonotalwaysreducetheamountofgreenwashundertakenbytheincumbentrm.Whengreenwashisjustsufcientlyhightoforeclosethemarketfromentrybythegreenrm,aggregateenvironmentalharmjumpsdownward.Thisdownwardjumparises,becausethetotalmarketquantityfallswhenentryisdeterred. 47

PAGE 48

Theglobalminimumforaggregateenvironmentalharmariseswhengreenwashisjustsufcienttoforeclosethemarketfromentrybythegreenerrm. Lemma3. Theglobalminimumforaggregateenvironmentalharmariseswhenthe(positive)amountofgreenwashisjustabovethelevelthatwouldforeclosethemarketbythegreenerrm.Anyincreaseinthexedentrycostmayservetoreduceaggregateenvironmentalharm.Fromtheperspectiveofreducingenvironmentalharm,marketregulatorsorenvironmentalactivistsmayprefergreenwashthatissufcientlyhightoforeclosethemarketfromentrybythegreenerrm.Asmentionedabove,thisregionislesspreferredbyconsumer-focusedregulators.Theglobalmaximumforconsumersurplusoccurswhenthemarketisnotforeclosed.Greenwashgenerallyservestodecreaseconsumerwelfareandincreaseaggregateenvironmentalharm.However,anumberofinterestingexceptionsmayaffectthedesirabilityofcurtailinggreenwash.Interestingly,consumerwelfareisnotmaximized,norisaggregateenvironmentalharmminimized,whengreenwashisperfectlyprevented.Understandingtheresponsivenessofagreenwashingrmtochangesinitsmarginalcostmayinformthosethatwishtocombat(orfacilitate)greenwash. 1.5ConcludingRemarksTheproceedingsectionsidentifyandaddressanexistinggapinthecurrentliteraturerelatingtofalseadvertisingforpollutingproducts.Thispaperanalyzesfalseadvertisingbehaviorbyabrownincumbentmonopolistinthepresenceofgreenerpotentialentrantwhenproducts`truepollutingpropertiescannotbereadilyobserved.Suchgreenwashadvertisingmakestheincumbent'sproductmoreappealingtoenvironmentallyconsciousconsumers,increasingitspotentialprots.Inthepresenceofpositiveentrycosts,greenwashadvertisingcanalsoaffectentrydecisionsbypotentialentrants,affectingtheoverallmarketstructure.Thispaperdemonstratesthatfalse 48

PAGE 49

advertisingmayenhanceconsumerwelfareandreduceaggregateenvironmentalharm,demonstratingthatsomepositiveamountofgreenwashadvertisingislikelydesirable.Greenwashdoesnotonlyservetoincreasetheincumbentrm'smarketshare;itcanalsoaltertheoverallmarketstructurebyforeclosingthemarketfromentry.Marketstructureeffectsmakeanalyzinggreenwashlessstraightforwardandaffecttheresultsconsiderably.Thereareinstancesinwhichincreasinggreenwashmaybenetconsumersand/ordecreaseaggregateenvironmentalharm.Infact,consumerwelfareincreasesandenvironmentalharmdecreaseswithcertain(positive)amountsofgreenwash.Consumersbenetwhentheperceivedlevelofproductdifferentiationisreduced,becausepricecompetitionbecomesmoreintense.Moreover,aggregateenvironmentalharmfallswhengreenwashadvertisingcanreducethenumberofrmsthatarepresentinthemarket,aslessrmsimpliesanoverallreductioninthemarketquantityofthepollutingproduct.Giventhesestructural-driveneffects,marketregulatorsandenvironmentalactivistsmustcarefulconsiderthelikelyeffectsofregulatinggreenwashadvertising.Theymaynotalwaysseektoreducegreenwash.Combattinggreenwashoftenoccursbyincreasingitscosteitherthroughlegalenforcementbymarketregulatorsoractivismandadvertisingcampaignsbyenvironmentalgroups.14Effectivelyelicitingaparticularresponsebyagreenwashingrmisnotasstraightforwardashighercostsresultinlessundesirablegreenwash.Properlycombatting(orfacilitating)greenwashrequirespreciseknowledgeofthemarketstructureandtherolethatpotentialentrantsplay.Inthepresenceofagreenpotentialentrant,aggregateenvironmentalharmincreases.Greenproductsattractnewconsumerstothemarket,andtheenvironmentalimpactoftheresultingincreaseinproductionisnotoffsetbyconsumersswitchingtothecleanervariant.Consequently, 14Thelatteraffectstherm'scostofgreenwashingbymakingconsumerslessgullible. 49

PAGE 50

inthepresenceofapotentialentrant,entry-deterringgreenwashmaybewelcomedbyenvironmentalgroups.Asapointoffutureresearch,themodelcouldbeadaptedtoincludeaprobabilitythatgreenwashisnotsuccessful,orevenreducestheincumbentrm'sdemandthroughnegativereputationeffects.Thisaddition,however,wouldnotaffectthedesirabilityofacertainpositiveamountofgreenwashadvertising.Additionally,adynamicmodel,inwhichconsumerslearnaboutthegood'sharmpost-purchase,couldbeexamined.Thispaperdoesnotconsiderthiscaseofexperiencegoods;rather,itonlyconsidersproductswhoseenvironmentalharmremainsunobservedpost-consumption,orcredencegoods. 50

PAGE 51

CHAPTER2SPEARSANDDYLAN?CHARACTERIZINGTHEINCENTIVETOBUNDLEWHENMIXEDBUNDLINGISNOTFEASIBLE 2.1IntroductionMusicconcerts,likeanyothergood,arriveatthemarketplaceafternumerousstrategicdecisionsbytheirproducers.Onesuchstrategicdecision,orpractice,thatappearsubiquitousintheconcertarenaisbundling.Thiscommonlyobservedpricingstrategy,inwhichrmsofferseveraluniquecomponentproductsasasinglecombinedbundle,isinplaceeverytimethatmultipleactsareplacedonthesameticket(e.g.,openingacts,jointtours,andfestivals).Previousinquiryintothispervasivebusinesspracticefocusesontheincentivetobundletwo(ormore)products,butfailstoaddressthecompositionofanoptimalbundle.Decidingwhethertobundletwoexistingproductsisanimportantbusinessdecision,butacommonquestionintheconcertticketbusinessis:whichactsshouldbebundled?Thispaperexaminestheincentivetobundle,butitsmoreimportantcontributionsurroundstheidenticationofanoptimalbundle.Musicconcertsdemonstratetheimportanceofidentifyinganoptimalbundle.Concertsareroutinelybundledintomultipleactevents,anditstandstoreasonthatthecompositionofthesebundlesisnotrandom.Doeventpromotersbundlesimilaractstoincreaseconsumers'willingnessestopay,ordotheybundlemoredifferentiatedeventstodrawalargercrowd?1Thispaperanswersthatquestionanddemonstrateshowconsumerspreferencesandcapacityconstraintsmayaffectthecompositionoftheprot-maximizingbundle.Bundlingcantakemanydifferentpracticalforms,buttwobroadcategoriesofbundlingstrategieshavebeenidentied.Bundlesareclassiedaseitherpurebundlesormixedbundles.Purebundlesconsistofcomponentproductsthatarenotavailable 1ConsistentwithKrueger(2005),Iconsiderpromoterstobetheprotmaximizers.Theanalysis,however,doesnotdependonselectingaparticularprotmaximizer. 51

PAGE 52

forseparatepurchase.Mixedbundlesconsistofcomponentsthatcanbepurchasedasadiscountedbundle,orseparately.Muchofthebundlingliteraturefocusesonmixedbundlinganddemonstratesthatmixedbundlingispreferred(byamonopolist)topurebundlingformanydifferentproducttypes.Mixedbundlingisaneffectivepricediscriminatingstrategy.Thispractice,however,israrelyobservedinconcertticketpricingexceptforthecaseofclassicalmusicseries.Musicconcertsaretypicallyofferedasabundleorseparately;thereisnodiscountofferedtoconsumerswhopurchaseatickettomorethanone.2Mixedbundlingmaybeundesirableinthisparticularmarketfortworeasons.First,ifweareconsideringsequentialactsinasinglevenue,itmaybetoocostlyorinfeasibleoreveninconvenienttoconsumerstoclearoutthevenuebetweenactsandre-checktickets.Thisexplanationisconsistentwithobservingmixedbundlinginclassicalmusicseries,whereactsplayonseparateevenings.Anotherreasonthatmixedbundlingisnotregularlyobservedinconcertticketsalesmaybetheexistenceofbindingcapacityconstraints.Whencapacityisconstrained,thereisnoneedtoofferticketstobundledcomponenteventsseparately,becausetheconsumerswiththehighestwillingnessestopayarealreadyllingalloftheavailableseats.Bothpracticalfeasibilityandbindingcapacityconstraintsmayexplainthelackofmixedbundlinginconcertticketsales.Thispaperexplorestheincentivetobundleeventswhenmixedbundlingisnotfeasibleandexaminesthedeterminantsofanoptimalbundle.Theremainderofthepaperisorganizedasfollows.Section2presentsasurveyoftheextantliterature.Section3developsamodelofhorizontallydifferentiatedproducts,inwhichconsumersexhibitconcavetransportationcosts.Itconsidersamonopolist'soptimalpricingand 2IacknowledgethatadiscountcouldbeofferedforconsumersthatattendoneconcertonFridaynightandanotheronSaturdaynight,butthispracticeisrarelyobservedoutsideofclassicalmusicseries. 52

PAGE 53

bundlingstrategieswhenthecomponentproductslocationsareexogenouslyassigned,andthen,considersthesameincentiveswhenthecomponentproductslocationsareendogenouslyselected.Idemonstratethattheincentivestobundleproductsaredependentbothonthelevelofdifferentiationbetweenthecomponentsandonthedegreeofadditivityofconsumersutilityfunctions.Lastly,Iexaminetheeffectofcapacityconstraintsontheincentivetobundle,ndingthattheuniquestructureofaparticularconstraintdetermineswhetheritreducesorintensiestheincentivetobundle. 2.2ExtantLiteratureMuchofthepreviousliteratureonbundlingincentivesfocusesontherelationshipbetweentheproducts,ormorespecically,ontherelationshipbetweenconsumerswillingnessestopayforthem.EarlyworkbyAdamsandYellen(1976)andSchmalensee(1984)demonstratethatamonopolistsincentivetobundleisstrongerwhenconsumerswillingnessestopayforthetwocomponentproductsareinverselyrelated.Schmalensee(1982)alsoshowsthatmixedbundlingisabetterpricediscriminationmechanismthanpurebundling.Whenconsumersvaluationsfortwoproductsareinverselyrelated,themonopolistcanincentivizemoreconsumerstopurchasebothproductsbybundlingthemintoasingleproduct.Moreover,byofferingadiscountedbundle,alongsideseparatelysoldcomponentproducts,themonopolistcanenjoythepricediscriminatingeffectofbundlingwithoutlosingconsumerswhosenetvaluationforoneofthecomponentsispositive,butwhosenetvaluationofthebundleisnegative.Theincentivetobundlecanalsobegreatlyaffectedbywhethertheproductsarecomplements,substitutes,orunrelatedgoods.Reservationvaluesforcomplementsaresuperadditive,whilereservationvaluesforsubstitutesaresubadditive,orinverselyrelated.Guiltinan(1987)showsthatrmsaremorelikelyto(purely)bundlecomplementarygoods,becausetheirreservationvaluesarestrictlysuperadditive.VankasteshandKamakura(2003),however,presentmorenuancedresults.Theydemonstratethatpurebundlingisoptimalforstrongcomplements,andsellingcomponentsseparatelyis 53

PAGE 54

optimalwithstrongsubstitutes.Theyndthatpurebundlingisstrictlysuboptimal(whenalsoconsideringmixedbundling)forsubstitutablegoods.Eachoftheabovepapersdescribescircumstancesunderwhichpurebundles,mixedbundles,orindividuallysoldcomponentsareoptimalstrategiesforthemonopolist,butnoneconsidertheconstructionofanoptimalbundle.Somemorerecentmarketingresearch(See,forexample,BradlowandRao(2000)andChungandRao(2003))hasattemptedtoapplyproductdesignandattributetheorytobundledesign,buttheirapproacheshavedifcultymeasuringthedegreeofheterogeneitybetweenbundlecomponentsandcharacterizingthesubstitutabilitybetweencomponents.Thecurrentpaperexaminestheincentivetobundlehorizontallydifferentiatedcomponents,identiesanoptimalbundledesign,anddescribestheroleofbothproductdifferentiationandadditivityofconsumers'utilityindeterminingtheseincentives.Thispapermakesanumberofmeaningfulcontributionstothetheoryofhorizontallydifferentiatedproductsandbundling.First,themodelexaminesboththeincentivetobundletwocomponentproductsandthedesignofanoptimalbundle.Italsofocusesonproductsthatcannotbesoldasamixedbundle,suchasadmissiontoevents.Asmoststudiespointtothedesirabilityofmixedbundlingastheprot-maximizingstrategy,thisapproachshineslightonagroupofproductsthatareoftenneglectedbythebundlingliterature.Thispaperalsoconsidersthepotentialformulti-purchasingandconcavetransportcostsinaHotellingmodel,whichhaveunfortunatelybeenunderstudied.Section3continueswiththeconstructionananalysisofthemodel. 2.3TheModelThemodelexaminestheincentivetobundlehorizontallydifferentiatedproductswhenmixedbundlingisnotfeasible.Whenbundlingisdesirable,andacontinuumofcomponentsisavailable,themodelidentiestheseller'sprotmaximizingbundlingstrategy.Theanalysisinthissectionispresentedinthecontextofconcertbundling,butitgeneralizestoanybundlingsituationinwhichmixedbundlingisnotpossible.In 54

PAGE 55

particular,itgeneralizestoanyscenarioinwhichadmissionfeesarechargedforentryintoaneventsuchaslmfestivalsandwinetastings.Themarketischaracterizedbyamonopolistsellerwhocanchoosebetweenhostingandsellingadmissiontotwoseparateevents,E1andE2,orhostingasinglecombinedevent,E12.ThecombinedeventE12issimplyabundleofcomponentactsE1andE2.Therearenomarginalcostsofproductionincludedinthemodel.Forentertainmentevents(e.g.,concertsorlmfestivals),zeromarginalcostisareasonableassumption.Fixedcostsarenotexplicitlytreatedinthemodel,butastheyarelikelytobehigherwhenthetwoeventsareofferedseparately,includingthemwouldonlyintensifytheincentivestobundlethatareexaminedthroughouttheremainderofthepaper.Thesellermaximizesprotsbychoosingwhethertobundlethetwoevents,selectingthe(horizontal)locationofthetwoevents,andchoosingthepricefortheproduct(s)offeredsimultaneously.Analytically,thedecision,orincentive,tobundlethetwoeventsisdeterminedbywhichoptionoffersthelargestprots,giventhatthelocationsoftheeventsandtheadmissionpriceswouldmaximizetheseller'sprotsineithercase.Itis,thus,irrelevanttoconsiderwhetherthebundlingchoiceismadeexante.Thefocusoftheproceedingsectionistoidentifywhetherthemonopolisthasanincentivetobundleandifso,whatcomponenteventsitshouldoptimallybundle.Consumersdifferintheirhorizontalpreferencesoverthepotentialevents.Forexample,inthecontextofmusicconcerts,someconsumerspreferrockmusicwhileotherspreferfolkmusic.Consumers'preferencesoverthepotentialeventsareuniformlydistributedalongaHotellingline,andtheutilityderivedbyaconsumerwithpreferencelocationifromattendingasingle-componenteventEjlocatedatljisgivenas: Ui,j=V)-222(jlj)]TJ /F6 11.955 Tf 11.95 0 Td[(ij.5,(2) 55

PAGE 56

whileaconsumerwithpreferencelocationithatconsumesbotheventsE1andE2,eitherasabundleorseparatelyderivesutility: Ui=(1+)V)-222(jl1)]TJ /F6 11.955 Tf 11.95 0 Td[(ij.5)-221(jl2)]TJ /F6 11.955 Tf 11.96 0 Td[(ij.5.(2)whereVisthegrossvaluationthataconsumerreceivesfromattendinganysingle-componentevent,and2[0,1]measuresthedesirabilityofattendingasecondevent.Theaboveutilityfunctionassumesthattheconsumers'grossvaluationofattendingmultipleeventsisweaklysubadditive.Thatis,thegrossvaluationofattendingasecondeventiseitherlessthanorequaltothatofthesecond.Thisspecicationisadoptedtocapturethenotionthataportionoftheutilityderivedfromaattendingaparticulareventisnotlocation-specicconsumersmaysimplyenjoyattendinganeventandthatportionofthegrossvaluationislikelydeclininginthenumberofeventsaconsumerattends.Thedisutilitythataconsumerreceivesfromattendinganon-idealevent(i.e.,aneventthatisnotperfectlyalignedwiththeconsumer'spreferences)isassumedtobeconcaveinthedifferencebetweentheconsumer'spreferencesandthehorizontallocationoftheevent.Theconcavityoftransportcostsassumptionissupportedbytworelated,practicalconsiderations.3First,consumersaremoreawareofdifferencesbetweeneventsthatareclosertotheirownpreferences.AfanofAmericanfolkmusicismorelikelytodistinguishbetweenaBobDylanandaWillieNelsonconcertthanbetweenaMadonnaandKylieMinogue.Fansof1980spopmusicwouldlikelymakethereversedistinction.Second,whentwodifferentiatedproductsarebundled,concavityoftransportcostsimpliesthatconsumerswhoseidealcomponentformspartofthebundleismorelikely 3Thedisutilitythataconsumerreceivesfromattendinganon-idealeventisreferredtoastransportcostsintheliteraturerelatedtohorizontalproductdifferentiationalaHotelling.See,forexample,d'Aspremont,Gabszewicz,andThisse(1979),Economides(1986),andCremer,Marchand,andThisse(1991). 56

PAGE 57

toconsumetheproductthanaconsumerwhoseidealcomponentislocatedbetweenthetwocomponentsthatformthebundledevent.Thatis,withconcavetransportcosts,anavidfanofonecomponentactismorelikelytoattendabundledconcertthanamoderatefanofbothacts.Itisalsoworthnotingthatthereisnoincentivetobundleinthepresenceofconvextransportcosts,makingsuchanassumptionirrelevanttothecurrentanalysis.Withconvextransportcosts,amonopolistalwaysearnsmoreprotbyofferingcomponenteventsseparately.Bundlingcomponentsneverincreasesdemand.Thoroughlycharacterizingtherm'soptimalpriceandbundlingstrategiesrequiresaninvestigationunderbothexogenousandendogenouslocationassumptions.Thesetwoassumptionsaddressdifferentpracticalsituations;neitherismorerealisticthantheother.Underanassumptionofexogenouslocations,Iinvestigatebundlingchoiceswhentheproductofferingisalreadydetermined.Forexample,awinerythatproducesPinotNoirandCabernetFrancfacesthechoiceofsellingtastingsforeachseparately(asiscommonatCanadianwineries)orasaight(asismorecommonatCaliforniawineries).Similarly,anentertainmentrmthatmanagestwoactswhoseconcerttourswillpassTampa,Floridaonthesameweekendmustchoosewhethertohostseparateeventsorcombinetheshowsintoabundledevent.Underbothscenarios,thelevelofhorizontaldifferentiationbetweenthetwoproductshasalreadybeenselectedwhenthepriceandbundlingstrategiesareselected.Underanassumptionofendogenouslocations,Iconsiderbundlingdecisionswhentheproductsvariantscanbeselectedfromapoolofpotentialproducts.Forinstance,thewinerymayhavealargenumberofdifferentvarietals(orevendifferentofferingswithinasinglevarietal)fromwhichtodesignitstasting.Itwouldbeabletochoosethelevelofdifferentiationbetweenthetwowinesthatisoffersandwhethertopricethemseparatelyortogether.Similarly,theentertainmentrmmaymanagemanyactsandbeabletoselectwhichactstoofferinTampa.Theabilityofthermtoselectthelevelofdifferentiationbetweenits 57

PAGE 58

productofferingsdeterminestheapplicabilityofanexogenousorendogenouslocationsassumption.BelowIconsiderpriceandbundlingstrategiesunderbothassumptions.ExogenousProductLocations:ComponentsSoldSeparatelyIdentifyingthemonopolist'sincentivetobundledifferentiatedproductsrequiresacharacterizationoftheprotsassociatedwithsellingthecomponentsseparately.Bundlingisonlydesirableifityieldshigherprotsthansellingthecomponentsseparately.Thissubsectiondescribesthemonopolist'sprot-maximizingpricingstrategywhenitdoesnotbundleundertheassumptionofexogenousproductlocations.ImodifythetraditionalHotellingmodeltoallowforthepotentialofmulti-purchasingbyasubsetofconsumers,identifyingthoseconsumerswhowillpurchasebotheventsevenwhentheyarepricedandsoldseparately.Tobegin,Iidentifythedemandforeachunbundledcomponent.Thedemandforaparticularcomponenteventisdependentonthelocationofbothcomponentevents,becauseconsumerscanengageinmulti-purchasing.Ifthecomponentsarenotsufcientlydifferent,someconsumerschoosetoattendbothevents.Whenthecomponentsaresufcientlydifferentiated,however,noconsumerchoosestoattendbothevents.Iftheeventsaresufcientlysimilar,someconsumerswillattendbotheventsevenwhentheyareofferedseparately.Figure2-1depictsthedemandrelationinthepresenceofmulti-purchasing.Dependingonthelevelofdifferentiationbetweenthetwoevents,itispossiblethatthoseconsumersforwhomoneofthecomponentsrepresentstheiridealeventengageinmulti-purchasing(Panel1a),oronlyagroupofconsumerswhoseidealeventliesbetweenthetwocomponenteventsengageinmulti-purchasing(Panel1b).Inbothcases,theconsumerwhosepreferencesplacethematthemidpointofallpurchasingconsumersarethemostlikelytoengageinmulti-purchasing.FollowingtheintuitionlaidoutbyAndersonet.al.(2010),multi-purchasinginaHotellingsettingariseswhenaconsumermayreceivepositiveutilityfromconsumingasecondvariant 58

PAGE 59

aftertheyhavepurchasedtheirmostpreferredvariant.Theindependentchoicestopurchaseeachvariantareonlylinkedthroughthedegreeofadditivityofutilities.Whenpreferencesarestrictlysubadditive(i.e.,<1),thechoicetopurchasethesecondvariantisonlyaffectedbyconsumptionoftherstvariantthroughthereductioninthegrossvaluation(i.e.,fromVtoV).4Inthecontextofattendinganevent,thisimpliesthatifaconsumerhasavaluationofattendinganevent,andifthatvaluationisdecreasinginthenumberofeventstheyattend,thersteventthataconsumerchoosestoattendonlyaffectsconsumptionofthesecondbyreducingtheconsumer'sinterestinattendinganevent.Prices,however,donothaveacompetitiveeffectbetweentheevents. Figure2-1. Multi-purchasing 4Here,secondmeanslesspreferred,andrstmeansmorepreferred.Itisnotnecessarilylinkedtotheorderinwhichtheproductsareconsumed. 59

PAGE 60

Thepotentialformulti-purchasinggivesrisetoadiscontinuityinthedemandfunctionwhenthepreferencesacrossthetwocomponentsarestrictlysubadditive.Whenmulti-purchasingoccurs,someconsumersofeacheventhavethereducedgrossvaluationofattendingthatevent,becauseitistheirsecondevent.Asaresult,thedemandforcomponenteventEiisgivenas: Di(pi,pj,d)=8>><>>:(V)]TJ /F7 11.955 Tf 11.95 0 Td[(pi)2+(V)]TJ /F7 11.955 Tf 11.96 0 Td[(pi)2,ifd<(V)]TJ /F7 11.955 Tf 11.95 0 Td[(pi)2+(V)]TJ /F7 11.955 Tf 11.95 0 Td[(pj)22(V)]TJ /F7 11.955 Tf 11.95 0 Td[(pi)2,ifd(V)]TJ /F7 11.955 Tf 11.96 0 Td[(pi)2+(V)]TJ /F7 11.955 Tf 11.96 0 Td[(pj)2(2)wheredisthedegreeofhorizontaldifferentiationbetweenE1andE2(i.e.,dl2)]TJ /F7 11.955 Tf 12.16 0 Td[(l1).5Whenthecomponentsaresoldseparately,cross-priceeffectsonlyarisethroughtheeffectofpricingonmulti-purchasing.Ifthepriceofoneeventincreases,itbecomeslesslikelythatanyconsumerchoosestopurchasebothproducts.Increasingthedegreeofhorizontaldifferentiationbetweenthetwoeventshasasimilareffect.Additionally,iftherearesomeconsumersthatdoattendbothevents,decreasingbelow1reducesthedemandforbothevents.Thatis,wheneverthetwoeventssharesomeattendees,andthegrossvaluationofattendinganeventisstrictlysubadditive,demandforeacheventisdecreasingin.Inordertocharacterizethemonopolist'spricingstrategy,Imustconsidertheirbehaviorbothwhenthelevelofdifferentiationbetweenthetwoeventsislowenoughtofacilitatemulti-purchasingandwhenitisnot.Ipayparticularattentiontothemonopolist'spricingbehavioraroundthemulti-purchasingthreshold.Multi-purchasingonlyariseswhen1)thecomponenteventsaresufcientlysimilar,and2)consumers'valuationofattendingasecondeventissufcientlylarge.Asproductsbecomemoredifferentiated,andasconsumersderivelessutilityfrom 5Forsimplicity,assumeE2islocatedtotherightofE1ontheHotellingline.Thatis,l2)]TJ /F7 11.955 Tf 11.95 0 Td[(l1>0. 60

PAGE 61

attendingasecondevent,themonopolists'abilitytosellbotheventstoasubsetofconsumersbecomeslesslikely.Theeffectofdifferentiationisstraightforwardandfollowsdirectlyfromtheconditionalstatementsinthepiece-wisedemandfunction.Thepricingstrategy,conditionalonthelevelofhorizontaldifferentiationbetweenthetwoevents,ischaracterizedasfollows: pi=8>><>>:2+2)]TJ /F2 11.955 Tf 6.59 8.17 Td[(p 2(4)]TJ /F10 7.97 Tf 6.59 0 Td[(1)]TJ /F15 7.97 Tf 6.59 0 Td[(2) 6V,ifd<()]TJ /F10 7.97 Tf 13.15 4.71 Td[(1 3)2V2V 3,ifd()]TJ /F10 7.97 Tf 13.15 4.71 Td[(1 3)2V2(2)Wecanseethatthemonopolistsetsitsper-componentpriceatpi=v 3henever1 3.Forsuchstronglysubadditivepreferences,themonopolistmaximizesprotsbyseparatingconsumersintotwodistinctmarketplaces.Noconsumersengageinmulti-purchasing,regardlessofthesimilaritybetweenthetwocomponentevents.Theeffectofontheprot-maximizingpricingstrategyinthepresenceofmulti-purchasingislessstraightforward,asitintroducesaninterestingtradeoffforthemonopolisttoevaluatewhenselectingitspricingstrategy.Themonopolistbenetsfrommulti-purchasingbehavior,asitprovidesthemonopolistwithalargerbaseofpotentialconsumers.It,however,alsofacestheconsequencethatthegrossvaluationofaneventisreducedformassofitscustomers.Thatis,thenumberofpotentialconsumersofeacheventisincreasing,buttheaveragegrosswillingnesstopayisfalling.Thistradeoffgivesrisetosomeaninterestingrelationshipbetweenthemarginalgrossvaluationofattendingasecondeventandtheprot-maximizingpricingstrategy.Figure2-2presentsthisrelationship.Itcanbeseeneasilythatpi=V 3for<1 3.Moreinterestingly,for21 3,2)]TJ 11.96 9.96 Td[(p 2,piisdecreasingin.Overthisrange,theprot-maximizingpricingstrategy,giventhatthemulti-purchasingconditionissatised,ischaracterizedbyfallingpricesasconsumers'willingnessestopayforasecondeventincrease.Whileagreaterdesiretoattendasecondeventincreasesthenumberofpotentialconsumersforbothevents 61

PAGE 62

Figure2-2. ComponentPricing throughmorecommonmulti-purchasing,theaveragevaluationofanattendeefalls,placingdownwardpressureonprices.Once>2)]TJ 12.25 9.95 Td[(p 2,however,furtherincreasesinthevaluationofattendingasecondeventplaceupwardpressureonthepriceofbothseparatelysoldevents.ExogenousProductLocations:BundledComponentsIdentifyingthemonopolist'sincentivetobundledifferentiatedproductsrequiresacharacterizationoftheoptimalbundlingstrategyandacomparisonoftheprotsassociatedwiththatstrategytotheprotsthatareearnedwhentheeventsaresoldseparately.Above,Iidentifythemonopolists'prot-maximizingbehaviorwheneventsarepricedandsoldseparately.Imustnowidentifytheoptimalpricingdecisionwhenthecomponenteventsareofferedasabundle.Icharacterizetheincentivetobundleasafunctionofboththedegreeofdifferentiationbetweenthecomponenteventsandthelevelofadditivityofconsumers'utilities. 62

PAGE 63

Tobegin,Icharacterizetherelationshipbetweenconsumers'preferencesforthebundledproductandthedegreeofhorizontaldifferentiationbetweenthecomponentevents.Thetwoconsumerswhoseidealeventsareincludedinthebundle(i.e.,theconsumerslocatedatl1andl2)receivethegreatestutilityfromconsumingthebundledproduct.6Thisimpliesthatdemandisonlypositivewhenthebundledprice,p12,anddegreeofdifferentiationaresuchthat(1+)V)]TJ 13.05 10.18 Td[(p d)]TJ /F7 11.955 Tf 13.06 0 Td[(p12>0.Additionally,oftheconsumerswithtasteslocatedbetweenl1andl2,theconsumerlocatedatl1+l2 2receivestheleastutilityfromconsumingthebundledproduct,implyingthatsubsetofconsumerslocatedbetweenl1andl2abstainfrompurchasingthebundledproductwhen(1+)V)]TJ /F8 11.955 Tf 11.07 0 Td[(2q d 2)]TJ /F7 11.955 Tf 11.07 0 Td[(p12>0.Asaconsequence,whenthepriceordegreeofdifferentiationincreasesaboveaparticularthreshold,someconsumerslocatedbetweenl1andl2choosenottopurchasethebundledproduct,butthoseconsumerslocatedatandnearl1andl2continuetopurchasethebundle.Asaresult,thedemandforthebundledeventisgivenbythefollowingpiece-wisefunction: D12(p12,d)=8>>>>>><>>>>>>:d2 2[(1+)V)]TJ /F9 7.97 Tf 6.59 0 Td[(p12]2+[(1+)V)]TJ /F9 7.97 Tf 6.59 0 Td[(p12]2 2,ifp12(1+)V)]TJ /F8 11.955 Tf 11.95 0 Td[(2q d 2d2 2[(1+)V)]TJ /F9 7.97 Tf 6.59 0 Td[(p12]2+[(1+)V)]TJ /F9 7.97 Tf 6.59 0 Td[(p12]2 2)]TJ /F8 11.955 Tf 9.3 0 Td[([(1+)V)]TJ /F7 11.955 Tf 11.95 0 Td[(p12]p 2d)]TJ /F8 11.955 Tf 11.96 0 Td[([(1+)V)]TJ /F7 11.955 Tf 11.96 0 Td[(p12]2,ifp12>(1+)V)]TJ /F8 11.955 Tf 11.96 0 Td[(2q d 2(2)Figure2-3providesagraphicalrepresentationoftherelationshipbetweenconsumers'tastesandcomponentcharacteristics.PanelAdepictsthecaseinwhichthecomponenteventsarenotsufcientlydifferentiatedtodeterconsumptionbyasetofconsumerslocatedbetweenl1andl2.PanelBdepictsthecaseinwhichthecomponenteventsaresufcientlydifferentiated.Bothpanelsgraphthelocationofthebundledproduct(L)andconsumers'individualpreferencesoverthehorizontallydifferentiatedproductattributes 6hisfeaturefollowsfromtheassumptionofconcavetransportcosts. 63

PAGE 64

inthel1,l2-plane.Consumersareuniformlydistributedalongthe45-degreeline,asthetastesofagivenconsumeri,i,donotdifferintheirrelationtothelocationsofthetwocomponentevents.Thoseconsumersthatchoosetopurchasethebundleareidentiedbythedarkenedportionofthe45-degreeline.Comparingthetwopanels,itisclearthatincreasingkabovezeroinitiallyincreasesdemand,butdecreasesaftercrossingacriticalthreshold. Figure2-3. BundleDemand Themonopolistmaximizesprotsbyselectingthebundledprice,p12,accordingto: maxp1212=p12D12(p12,d)(2)Themonopolistsetsthebundledpricesuchthattheconsumerwhosepreferencesarelocatedatthemidpointbetweenthetwocomponenteventlocationsreceivesnoconsumersurplus.Theprotfunctionisnon-differentiableatthekinkinthedemandfunction.Itis,however,monotonicallyincreasinginp12whenp12<(1+)V)]TJ /F8 11.955 Tf 12.54 0 Td[(2q d 2 64

PAGE 65

andmonotonicallydecreasinginp12whenp12>(1+)V)]TJ /F8 11.955 Tf 13.07 0 Td[(2q d 2,implyingthatp12=(1+)V)]TJ /F8 11.955 Tf 11.96 0 Td[(2q d 2maximizesthemonopolist'sprots.Ifthemonopolistchoosestobundlethetwocomponentevents,itearns12=5d 4h(1+)V)]TJ /F8 11.955 Tf 11.96 0 Td[(2q d 2i.Comparingthepotentialprotsassociatedwithbundlingthecomponenteventstothoseearnedwhentheeventsareofferedseparately,itcanbeseenthattheincentivetobundleisdependentonthedegreeofhorizontaldifferentiationbetweenthetwoeventsandthemarginalvaluationofattendingasecondevent.Thefollowingdiscussionregardingtheincentivestobundleispresentedasfollows.Forgivenrangesof,thedegreesofexogenouslydeterminedproductdifferentiationthatprovideanincentivetobundlewillbecharacterized.TheserelationshipsarealsopresentedinFigure2-4.<0.769:Forlowvaluesof,themonopolistalwaysearnsmoreprotssellingadmissiontothetwoeventsseparately.When,consumersdonotplaceahighvalueonattendingasecondevent,themonopolistisbetteroffsellingadmissionseparately,regardlessofthedegreeofproductdifferentiationbetweenthetwoevents.Interestingly,thisrangeofincludestherangeofdoverwhichtheprot-maximizingpriceofadmissiontoeacheventisdecreasingin.0.769<<0.857:Withinthisrangeof,asinglerangeofdoverwhichthemonopolisthasanincentivetobundletheeventsarises.Thisrangeisboundedtotheleftandright,suchthatwemaydeneitas[d ,d],andd>0.Forverysimilarevents(i.e.,d<[d ),themonopolistearnsgreaterprotsbypricingadmissionseparatelyandallowingasubsetofconsumerstoengageinmulti-purchasing.Forsufcientlydifferentiatedevents(i.e.,d>d),themonopolistearnsgreaterprotsbypricingadmissionseparatelyatp=V 3andeliminatingthepossibilityformulti-purchasing.0.857<<0.884:Withinthisrangeof,twouniquerangesofdoverwhichthemonopolisthasanincentivetobundletheeventsarise.Similartotherangesdescribeddirectlyabove,themonopolistisbetteroffpricingtwoeventsthatarecharacterized 65

PAGE 66

byverysmallandverylargedegreesofdifferentiationseparately.Withinthisrangeof,however,thereisalsoathirdrangeofdinwhichthemonopolistdoesnotchoosetobundlethetwoevents.Thisthirdrangearisesduetotheupwardjumpintheprot-maximizingpricingstrategy(whenthecomponenteventsaresoldseparately)whendissufcientlylargetopreventmulti-purchasing.Withinthisrangeof,thatupwardjumpislargeenoughtomakebundlinglocallyunprotable.>0.884:Withinthisrangeof,aswith0.769<<0.857,asinglerangeofdoverwhichthemonopolisthasanincentivetobundletheeventsarises.Thevalueofattendingasecondeventissufcientlyhightomakebundlingprotable,evenfordegreesofhorizontaldifferentiationthatwouldgiverisetomulti-purchasingintheabsenceofproductbundling.Whenconsumersdonotplaceahighvalueonattendingasecondevent(i.e.,isparticularlysmall),bundlingisnotprotable.Themonopolistearnshigherprotssellingthetwoproductsseparately.Onlywhen>0.769canbundlingbeprotable;themonopolistsellssufcientlysimilarorsufcientlydifferentcomponenteventsseparately,butoffersmoderatelydifferentiatedeventsasabundle.Theremainderofthissectionconsiderstheeffectofendogenizingthedegreeofproductdifferentiationbetweenthetwoevents.Thismodicationtothemodeladdressestheissueofconstructingtheoptimalbundle.Icharacterizetheoptimaldegreeofhorizontaldifferentiationasafunctionoftheadditivityofconsumers'utilities.EndogenousProductLocations:ComponentsSoldSeparatelyAswasdemonstratedabove,themonopolists'protsarereducedwhentheproductsaresufcientlysimilartogiverisetomulti-purchasingbyasubsetofconsumers.Themonopolistcanavoidsufferingthesereducedprotswhenconsumers'preferencesarenotstrictlyadditivebyincreasingdifferentiation.When<1,themonopolistenjoysanincreasedwillingnesstopayforitscomponenteventsifthedegreeofhorizontaldifferentiationbecomessufcientlystrongtodeteranyconsumerfrom 66

PAGE 67

Figure2-4. TheIncentivetoBundle purchasingbothproducts.7Thatis,themonopolistincreasesitsownprotsbysettingd>()]TJ /F10 7.97 Tf 11.89 4.7 Td[(1 3)2V2.Theprot-maximizationproblemforthemonopolistthatsellscomponenteventsseparatelyissimpliedto: maxpii=2pi(V)]TJ /F7 11.955 Tf 11.96 0 Td[(pi)2(2) 7Iexplainthattheprotsassociatedwiththecaseofstrictadditivityofpreferences(i.e.,=1)resultsinthesamelevelofprotswhentheeventsarenotbundledbelow. 67

PAGE 68

Thermsetsapriceofp1=p2=V 3foreachofitseventsandsellsD1=D2=8V2 9unitsofeachevent.Consequently,themonopolistearnstotalprots(acrossthetwoevents)equalto=16V3 27.Thislevelofprotsarisesforanydegreeofdifferentiationbetweenthetwoeventsaboveacriticalthresholdmaximizesprots.Protsareconstantforanydthatsatisesthisthreshold.Whenconsumers'preferencesacrossthetwoeventsarestrictlyadditive,theresultingprotsequalthoseobtainedabove.Inthisparticularcase,demandsimpliestoDi(pi)=2(V)]TJ /F7 11.955 Tf 12.27 0 Td[(pi)2,whichisidenticaltotheportionofthedemandcurveonwhichthemonopolistoperatesabove.Theonlydistinctionbetweenstrictsubadditivityandstrictadditivityofpreferencesisthatinthecaseofstrictadditivity,themonopolistdoesnothavetosatisfyaminimumlevelofdifferentiationbetweenthetwoevents.Anylevelofdifferentiationresultsinprotsequalto=16V3 27.EndogenousProductLocations:BundledComponentsIfthemonopolistchoosestobundlethetwoevents,regardlessofthedegreeofdifferentiationbetweenthecomponentevents,themonopolistmaximizesprotbysettingp12=(1+)V)]TJ /F8 11.955 Tf 12.63 0 Td[(2q d 2.Thus,themonopolist'maximizationproblemcanbestatedas: maxd12=5d 4"(1+)V)]TJ /F8 11.955 Tf 11.96 0 Td[(2r d 2#(2)Themonopolistmaximizesprotbysettingd=2(1+)V 32andp12=(1+)V 3andearnsaprotof12=10 4(1+)V 33.Comparingthemaximumprotsassociatedwithbundlingandthoseassociatedwithsellingthecomponenteventsseparately,theincentivetobundlebecomesquitesimple.Ifthemonopolistisabletoselectthedegreeofdifferentiationbetweenthetwocomponentevents,itchoosestobundlethetwoeventswhenconsumerssufcientlyvalueattendingasecondevent,when>0.857.Thatis,themaximumprotsthatitcanearnwhenbundlingthetwoeventsexceedsthoseitcanearnwhentheeventsarepriceseparatelyifconsumersexhibitastrongdesiretoattendasecondevent. 68

PAGE 69

Interestingly,whenthemonopolistchoosestobundlethetwoevents,itsetsadegreeofdifferentiationgreaterthanthatwhichitmightselectifitsellsthecomponenteventsseparately.Theprotmaximizingdegreeofdifferentiationwhentheeventsarebundledisgreaterthanthethresholdlevelthatwasestablishedabovewhenconsideringseparatelypricedevents.Thisresultimpliesthat,dependingonthevaluesofconsumers'grossvaluationofattendingasecondevent,itispossibletoseetwoeventsthataremorecloselyalignedofferedseparately,whilesomemoredifferentiatedonescouldbeofferedasabundle. Proposition2.1. Whenthedegreeofdifferentiation(i.e.,horizontallocation)isen-dogenouslyselected,themonopolistoffersmoredifferentiatedcomponentswhencomponentsarebundledthanitwouldifcomponentsaresoldseparately.Alternativelystated,amonopolistthatbundlesoffersmoreproductdiversitythanwouldbeofferedwithinaduopolymarket.Inthissection,Ihavecharacterizedamonopolist'spricingandbundlingincentiveswhenitofferstwohorizontallydifferentiatedproductsandfacesconsumerswithconcavetransportcosts.Bundlingonlyarisesasaprotmaximizingstrategywhenconsumersderiveasufcientlylargeamountofutilityfromattendingasecondevent,andthatthresholdlevelismoststrictwhenthemonopolistendogenouslychoosesthedegreeofdifferentiationbetweenthetwoproducts.Whenthedegreeofproductdifferentiationisexogenouslydetermined,bundlingisneverobservedwhen<0.769.Whenconsumers'marginalgrossvaluationofattendingasecondeventrisesabovethisthreshold,themonopolistchoosestobundlemoderatelydifferentiatedproducts.Verysimilarandverydissimilarproductsareofferedseparately.Whenthedegreeofdifferentiationbetweenthetwoproductsisendogenouslychosenbythemonopolist,bundlingonlyarisesif>0.857.Interestingly,whenthisthresholdisnotmet,themonopolistmaychoosetooffertwoeventsseparatelythatarelessdifferentiatedthanthebundlethatisofferedwhenthethresholdismet. 69

PAGE 70

2.4Extension:CapacityConstraintsWhenconsideringeventsastheproductsbeingbundled,alogicalextensionistoconsidertheeffectofcapacityconstraints.Courty(2003)claimsthat39%ofallshowssellout,implyingthatabindingcapacityconstrainthasbeenmetunless,ofcourse,theprot-maximizingquantitysuppliedjustequalsthesizeoftheconstraint.Intheprecedingsection,bundlingallowedthemonopolisttoextractadditionalsurplusfromeachconsumerandincreasedtheoverallnumberofconsumerspurchasing.Anaturalquestiontoaskiswhetherlimitingtheextenttowhichbundlingcanbeusedtoattractadditionalconsumersdiminishestheincentivetobundle.Thissectiondemonstratesthatthatisnotthecase;tothecontrary,capacityconstraintsmay,infact,intensifytheincentivetobundle.Capacityconstraintscanbeappliedtoeventsintwodistinctways:1)toeachproduct(separately),or2)totheoverallmarketquantity(jointly).Therstmethodofapplyingacapacityconstraintwouldentailconstrainingeachunbundledcomponentatthesamelevelasthebundledproduct,whereasthesecondwouldentailconstrainingthesumoftheunbundledcomponentsatthesamelevelasthebundledproduct.Anybindingconstraintonquantity,regardlessofhowitisappliedtounbundledcomponents,affectsthepricingandcompositionoftheprot-maximizingbundle.Themonopoliststillsetsthepriceanddegreeofdifferentiationsuchthattheconsumerlocatedatl1+l2 2isindifferentbetweenpurchasingthebundleandnotpurchasingthebundle(i.e.,thatconsumerreceiveszeroconsumersurplus).Ifthecapacityconstraintissetatk,thisimpliesthattheprot-maximizingdegreeofdifferentiationisd*=4k/5,andpriceisp12=(1+)V)]TJ /F8 11.955 Tf 11.98 0 Td[(2q 2k 5.Whenthecapacityconstraintbinds,theprotsassociatedwithbundlingaregivenask[(1+)V)]TJ /F8 11.955 Tf 12.01 0 Td[(2q 2k 5].Asaresult,tighteningthequantityrestrictiondecreasesthedegreeofdifferentiationandincreasesthepriceofthebundledproduct.Asthetotaloutputassociatedwithbundlingisgreaterthanthatofunbundledcomponents,capacityconstraintsinitiallyaffectonlytheprotsassociatedwithbundling. 70

PAGE 71

Capacityconstraintsthatonlybindwhenthecomponenteventsarebundledmustreducetheincentivetobundle.Itislessobviouswhathappenswhentheconstraintbindsforunbundledcomponentsaswell.Byadoptingoneofthetypesofquantityrestrictionsmentionedabove,itispossibletofullycharacterizetheeffectofcapacityconstraintsontheincentivetobundleincludingwhencapacityconstraintsaffectthepricingofunbundledcomponents.CapacityConstraintsImposedonUnbundledComponentsSeparatelyCapacityconstraintsmaybeimposedsuchthateachindividualcomponentfacesthesameconstrainedquantityasthebundle.Thistypeofquantityconstraintislikelytoarisewithmusicconcerts,inwhichtheconstraintisimposedbythesizeofthevenue.Iftwo(bundled)actsarescheduledattheTampaTheatreonFridaynight,themaximumquantityofticketsthatmaybesoldisconstrainedbythenumberofseats.Ifthetwoactswereofferedasunbundledcomponents,withoneactplayingFridaynightandoneplayingSaturdaynight,eacheventisindependentlyconstrainedbythenumberofseatsinthetheatre.Wheneachactisofferedseparately,eachmaytakeadvantageofthefullsizeofthevenue;thequantitysoldofthebundledeventcouldnotexceedthatsameoccupancylimit.Forcapacityconstraintsthataresufcientlylargetobindonlythebundledquantity,theincentivetobundleismaintainedif:k[(1+)V)]TJ /F8 11.955 Tf 11.96 0 Td[(2q 2k 5]>16V3 27.Whentheconstraintisrestrictiveenoughtobindthequantityofunbundledcomponentssoldaswell(i.e.,k<8V2 9),themonopolistchoosestobundleif: maxdB)]TJ /F6 11.955 Tf 11.95 0 Td[(NB=k"(1+)V)]TJ /F8 11.955 Tf 11.96 0 Td[(2r 2k 5)]TJ /F8 11.955 Tf 11.96 0 Td[(2V+2r k 2#>0.(2) 71

PAGE 72

TherelationshipbetweenthequantityrestraintandtheadditivityofconsumersutilityisdepictedinFigure2-5.8Theshadedregioniscomprisedofallcombinationsofand^kthatmaintaintheincentivetobundle.Twointerestingpointsariseoutofthisanalysis.First,thereisnobindingconstraintthatthatdoesnotreducethebundlingincentive,relativetotheunconstrainedequilibrium.Theminimumlevelofadditivitythatmaintainstheincentivetobundleis>0.859,comparedto>0.857intheunconstrainedequilibrium.Second,tighteningtherestraintbeyondk=8V2 9alwaysreducestheincentivetobundle.Thatis,maintainingtheincentivetobundlerequiresahighervalueof.Thisresultisratherintuitive.Ifthemonopolistisabletoofferthetwoeventsastwoseparateproducts(unbundledcomponents)andeasetheeffectofthecapacityconstraint,tighteningthequantityrestraintmakesitmorelikelytodoso.CapacityConstraintsImposedontheSumofUnbundledComponentsItisalsoreasonabletoconsiderhowtheincentivetobundleisaffectedwhenthecapacityconstraintbindsthesumofoutput(acrossthetwounbundledcomponents).Constraintsmayapplyinthiswasformulti-venueormulti-stageevents.Suppose,forexample,thattheGasparillaMusicFestivalinDowntownTampacansellticketstoeventsontwodifferentstagesseparatelyorjointly.Assumingthatconsumerswillself-separatetolleachvenue,thequantityconstraintwouldapplynottotheindividualstages,butthetotalacrossthetwo.Regardlessofwhetheradmissiontothetwoparksissoldasabundleorunbundledcomponents,aquantityconstraintwouldapplytothesumofproductssold.Whentheoverallmarketquantityofunbundledproductsisconstrainedatthesamelevelofthebundle,andtheconstraintisrestrictiveenoughtobindthequantityof 8Inordertosimplifythegraphicalrepresentation,Ihaveplacedanadjustedquantityconstraintonthehorizontalaxis,where^kk V2. 72

PAGE 73

Figure2-5. CapacityConstraintsImposedOnUnbundledComponentsSeparately unbundledcomponentssold(i.e.,k<16V2 9),themonopolistchoosestobundleif: maxdB)]TJ /F6 11.955 Tf 11.96 0 Td[(NB=k"(1+)V)]TJ /F8 11.955 Tf 11.95 0 Td[(2r 2k 5)]TJ /F7 11.955 Tf 11.95 0 Td[(V+r k 4#>0.(2)TherelationshipbetweenthequantityrestraintandtheadditivityofconsumersutilityisdepictedinFigure2-6.Theshadedregioniscomprisedofallcombinationsofand^kthatmaintaintheincentivetobundle.Unlikethecaseofcapacityconstraintsthatbindunbundledcomponentsindependentlydescribedabove,therearesomevaluesforthecapacityconstraintthatintensifytheincentivetobundle,relativetotheunconstrainedequilibrium.Capacityconstraintsthatapplytothesumofunbundledcomponentscanmakebundlingmorelikely.Additionally,tighteningtheconstraintintensiestheincentivetobundle. 73

PAGE 74

Figure2-6. CapacityConstraintsImposedOnUnbundledComponentsJointly Figure2-7comparesthetwotypesofquantityrestrictionsandtheireffectsontheincentivetobundle.ItoverlaysFigures2-5and2-6.Itisclearthattheincentivetobundleisalwaysstrongerwithconstraintsthatapplytounbundledcomponentsjointly.Thisresultisintuitive;themonopolistisnotabletoavoidtheconstraintbysplittingthebundleinthiscase.Capacityconstraintsalwaysweakentheincentivetobundlewhentheyonlybindonthe(higher)bundledquantity.Thatis,whenaquantityrestrictionthatonlyaffectsthebundleistightened,maintainingtheincentivetobundlerequiresahighervalueof.Astheconstrainedcapacityfallstothepointthatitaffectstheunbundledcomponents,however,tighteningthequantityrestrictiondoesnotalwaysreducetheincentiveto 74

PAGE 75

Figure2-7. CapacityConstraintsCompared bundle.Iftherestraintappliestounbundledcomponentsjointly,tighteningabindingconstraintincreasestheincentivetobundle. 2.5ConclusionTheprecedingsectionshaveexaminedtheincentivetobundleeventsfocusingonthespeciccaseofmusicconcertsandidentiedanoptimalbundle.Theextantliteraturefocusesprimarilyonfactorsaffectingtheincentivetobundlecomponents,butfailstocharacterizeanoptimalbundleadequately.Musicconcerts(andmoregenerally,admission-basedevents)oftenpresenttheinterestingquestionofwhatactsshouldbebundled.Additionally,theliteratureplacesmuchofitsattentiononmixedbundles,butthispracticeisnotwidelyobserved.Concertticketsaretypicallysoldasabundleorseparately. 75

PAGE 76

Thispaperaddresseswhatproductsamonopolistwouldchoosetobundle.Previousattempts(inthemarketingliterature)areincapableofsimultaneouslyinvestigatingdifferentiatedproductsandmaintainingacharacterizablerelationshipbetweenthoseproducts.Thatis,theydescribehowproductattributesshoulddiffer,buttheydonotdescribethedegreeofcomplementarityorsubstitutabilitybetweentheproductvariants.Componets,inthispaper,arehorizontallydifferentiated,andoptimalbundlesarecharacterizedbythedegreeofdifferentiationbetweenthecomponentevents.Thissettingalowsfortheidenticationoffactorsthataffectbothincentivetobundleandthecompositionoftheoptimalbundle(e.g.,theadditivityofconsumers'preferencesandcapacityconstraints).Thispapershowsthatboththeincentivetobundleandthedeterminationoftheoptimalbundledependuponthelevelofadditivityofconsumersutilities.Whenthemonopolistcannotselectthedegreeofdifferentiationbetweenthecomponentevents(i.e.,itcannotconstructanoptimalbundle),itchoosestoofferverysimilarandverydifferentcomponentsseparately,butbundlesmoderatelydifferentiatedproducts.Themonopolistbundlesdifferentiatedproductstoincreasethedemandforitsbundle,butitalsowantstolimitthedisutilitythatconsumerssufferfromattendinganactthatissubstantiallydifferentfromtheirindivdualtastes.Increasingthelevelofadditivityexhibitedbyconsumerspreferencestypicallyincreasestheincentivetobundle,andresultsinamoredifferentiatedoptimalbundle.Capacityconstraintsalsoaffecttheincentivetobundleandthecompositionoftheoptimalbundle.Whetheracapacityconstraintreducesorintensiestheincentivetobundledepandsonwhethertheconstraintappliestounbundledcomponentsseparatelyorjointly.Iftheconstraintappliesseparately,theincentivetobundleisreduced,becausethemonopolistcanofferthecomponentsseparatelyandexpandoutput.Iftheconstriantappliesjointly,theincentivetobundleisintensied. 76

PAGE 77

Lastly,thepaperdemonstratesthatduopolistsmaychoosetoofferlessdifferentiatedproductsthanamonopolistthatchoosestobundle.Theduopolistsdifferentiatetheireventsenoungtoseparatethemarketsinthesinglermsetting,thislevelofdifferentiationcorrespondstothecriticallevelatwhichmulti-purchasingisavoided.Themonopolist,however,wishestoincreasedifferentiationbetweenbundledcomponentsabovethislevelinordertoextractadditionalsurplus. 77

PAGE 78

CHAPTER3HOSPITALMERGERSANDPRODUCTDIFFERENTIATION:ANECONOMICAPPROACH 3.1IntroductionThehealthcareindustryhasexperiencedconsiderableconsolidationsincethemid-1990s,leadingtogreatermarketconcentration.Mergeractivityslowedbetween2002and2009(withapproximatelysixtymergersannouncedannually),buthasbeenontherisesince.Seventy-twoandninetymergerswereannouncedin2010and2011,respectively.MarketanalystsforecastthathealthcaremergerswillbecomemorecommonasthePatientProtectionandAffordableCareActisimplemented.Thenewhealthcarelawcreatesstrongerincentivesforhospitalstomerge.Thecurrentframeworkforevaluatingmergers,laidoutintheFederalTradeCommissions2010HorizontalMergerGuidelines,doesnotadequatelyaddresstheparticularitiesofthehealthcaremarket.Inparticular,itdoesnotassesstheeffectsofcombinationsonnon-price(e.g.,quality)competitioninthepresenceofsubstantialpriceregulation.AllmergingrmsinconcentratedindustriesmustnotifytheFederalTradeCommission(FTC),whichundertakesapreliminarymergerscreeninganddetermineswhethertochallengethemergerinfederalcourt.Proposedhealthcaremergersarenotiedthroughthesamemergerscreeningprogramasinotherindustriesandaresubjecttothereviewprocessoutlinedinthe2010HorizontalMergerGuidelines.Ifamergerisexpectedtoresultinsubstantiallyreducedcompetitionthatislikelytoharmconsumersthroughhigherprices,theFTCseeksapreliminaryinjunctionandmayultimatelychallengethemergerthroughatrialonthemerits.1 1OncetheFTCisawardedapreliminaryinjunction,mergingpartiestypicallyabandontheproposedmerger.Thus,iftheFTCisawardedthepreliminaryinjunction,themergeriseffectivelyblocked. 78

PAGE 79

The2010HorizontalMergerGuidelinesoutlinetheextensiveanalysisthatisconductedduringtheFTCsmergerscreeningprocessandprovidesinformationregardingwhatmergersarelikelytobechallenged.Theanalysisthatislaidoutintheguidelinesisintendedtodeterminewhetheraproposedmergerislikelytoincreasemarketconcentrationandresultinhigherpricespaidbyconsumers.CoateandUlrick(2006)demonstratethatthenumberofsignicantcompetitorsandtheincreaseintheHirschmanHerndahlIndex(HHI)arestrongpredictorsofwhetheraproposedmergerislikelytobechallengedbytheFTC.TheguidelinesandanalysesofmergerenforcementoftheFTC,however,arerelativelyquietregardingthelikelynon-priceeffectsofaconsummatedmerger;almostallattentionispaidtowhetheramergerisexpectedtoincreasepricesthrougheitherunilateralorcoordinatedeffects.Thisapproachdoesnotaddresswhetheraproposedmergerislikelytoreduceservicequality,whichisaparticularlyimportantquestioninprice-regulatedmarkets.Thispaperfocusesonhealthcaremarketsinwhichpricesaresetbyaregulatororgovernmentagency.Thisfocusisadoptedfortworeasons.First,anumberofhealthcaremarketsareeitherprice-regulatedoreffectivelyprice-regulated.IntheUnitedStates,healthcareprovidersthatservicelargenumbersofelderlyorlow-incomepatientsareprice-regulatedthroughtheMedicareandMedicaidprograms.Thedimensiononwhichrmscompeteinthesemarketsisservicequality,anditisworthwhiletoexaminetheuniqueeffectsofmergersinthesemarkets.Anotherjusticationforthisapproachisthedesiretoexaminequalityeffectsinisolation.Removingprice-basedstrategicbehavior,amergermaybeexaminedonlyforitseffectonstrategiceffectsrelatedtoquality.Ashealthcaremergersbecomemorecommon,andtheFTCisfacedwiththedecisionofwhethertochallengeoracceptincreasedmarketconsolidation,itisimportantthatpractitionersareinformedofthelikelyeffectsofhorizontalmergersinthehealthcareserviceindustry.Thispaperexaminesthelikelyeffectofmergerson 79

PAGE 80

servicequalitywhencompetitorsareprice-regulated.Theremainderofthispaperisorganizedasfollows.Section2describestheextensiveliteratureaddressingtheeffectofcompetitionandmarketconsolidationonqualityinhealthcaremarkets.Section3presentsasimplemodelofmergersinmarketscharacterizedbyverticalandhorizontalproductdifferentiation.Themodelisemployedtopredictpost-mergerqualityadjustmentsinthepresenceofpriceregulation.Section4providesconcludingremarks. 3.2HealthcareCombinationsandServiceQualityHealthcaremarketsaredifferentiatedproductsmarkets.Fromwaitingtimestospecializations,fromtheratioofnursepractitionerstodoctors,notwohealthcareprovidersarethesame.Consumerschoosetheirproviderbased,inpart,ontheproductattributesoftheavailablechoices.2Theattributestheyconsiderarebothhorizontalandvertical.Horizontalattributesincludespecializationsandgeographicalproximitytothepatient.Verticalattributesincludequalityoffacilities,doctor-to-patientratios,andwaitingtimes.Bothtypesofdifferentiationaffectconsumerswelfare.Asaconsequenceitisimportanttocharacterizetheeffectofmarketconsolidationonservicequalityandvariety.Thissectionhighlightssomeoftheimportantndingsinthisarea,examiningbothempiricalandtheoreticalwork.Competitionandregulationaffectthedegreeofdifferentiationinboththehorizontalandverticaldimensions.Thefocusofexistingempiricalanalysesfocusesontheverticaldimensionandprovidesmixedresultsregardingtheeffectofmarketconcentrationonquality.KesslerandMcClellan(2000),HoandHamilton(2000),Sari(2002),Tay(2003),andKesslerandGeppert(2005)ndthatincreasingmarketconcentrationreduces 2Thesetofavailableprovidersis,inpractice,oftendeterminedbytheinsurancecompanywithwhichtheconsumerhascoverage.Theanalysispresentedhere,however,isonlyconcernedwiththestrategicbehaviordrivenbyconsumerschoices. 80

PAGE 81

quality,whileMukamelet.al.(2001),andVolppet.al.(2005)identifytheoppositeeffect.Thislargebodyofexistingliteratureconcerningtheeffectofhealthcaremergersonservicequalityidentiesinconclusiveresults.Moreover,thereislittleempiricalresearchregardingtheeffectofhealthcaremergersonthevarietyofserviceandspecializations.Theextanttheoreticalliteraturetreatshealthcaremarketsasspatiallydifferentiatedproductsmarketsandinvestigatestheeffectofaddingadditionalcompetitorsonservicequalityandvariety.Brekkeet.al.(2008)examinetheeffectofaddingadditionalcompetitorsonwaitingtimesandndthatreducingtravelcostsincreaseswaitingtimes,andthathigherhospitaldensityreduceswaitingtimes.Thesecondresultismoreapplicabletoananalysisofmergers.Brekkeet.al.(2011)examinetheeffectofaddingadditionalcompetitorsonservicequalityandndthattheeffectofcompetitiononservicequalityisambiguous.Astheauthorspointout,servicequalityissimilarto(theinverseof)waitingtimes,buttheirlaterworktreatsqualityimprovementsascostlyinvestments,whereastheyhadpreviouslytreatedwaitingtimereductionsascostless.Bothanalysesassumethathealthcareprovidersarepriceregulated.Theauthorsdonotexaminetheeffectofmarketconcentrationonhorizontaldifferentiation.TheanalysesconductedbyBrekkeet.al.arelimitedbytheiromissionofprices.3Additionally,theydonotdirectlyaddresstheeffectofmergersonproductquality;theysimplyexaminewhetheraddinganadditionalcompetitorenhancesorreducesservicequality.Thetwoquestionsareparticularlydifferentinaspatialsetting.Mergersdonoteliminatecompetitorsinthissetting;theysimplyhandcontrolofonecompetitorovertoanother.NormanandPepal(2000)ldirectlyaddresstheimpactofmergers,buttheiranalysisassumeshomogenousservicequalityacrosscompetitors.Theyonlyinvestigatetheeffectofmergersonproductvariety,ndingthatmergersincreasevariety. 3WhileIdonotconsiderprice-settingbytheindividualhospitals,Iexaminethedifferentialeffectsbetweenstandardizedandindividualizedregulated-pricesettings. 81

PAGE 82

Empiricalanalysesidentifyconictingevidenceregardingtheeffectofmergersonservicequality.Theydonotaddresstheeffectofmergersonproductvariety.Theoreticalanalysesarelimited.Noexistingpaper,tomyknowledge,examinestheeffectofmergersonbothverticalandhorizontaldifferentiationinsettingscharacterizedbysubstantialpriceregulation. 3.3TheoreticalModelofVerticalandHorizontalDifferentiationThissectiondevelopsamodelofqualitycompetitionamonghorizontallyandverticallydifferentiatedhealthcareserviceprovidersinthepresenceofregulatedprices.Aregulatorannouncesprices;observingtheseprices,eachproviderchoosesitsownlevelofservicequality.Theanalysisconsiderstheuniquecasesofasingle,standardizedregulatedpricethatisappliedtoallrmsandindividualizedregulatedprices.Thesetwocasesarecompared.Thesectionproceedsbyidentifyingtheeffectsofamergerineachofthesesettings.Patientsareconcernedwithboththehorizontallyandverticallydifferentiatedattributesofdifferentserviceprovidersanddifferonlyintheirhorizontalpreferences.Patientsareuniformlydistributed(inhorizontalpreferenceorgeographicspace)alongtheinterval[0,1].4Eachpatientchoosestopurchaseoneunitofservicefromtheproviderwhoseserviceoffersthehighestlevelofnetutility.Apatientwithhorizontalpreferenceithatobtainsservicefromahealthcarefacilitywithspecializationljandqualityqjreceivesutility: U(pj,qj,i,lj)=V+qj)]TJ /F7 11.955 Tf 11.96 0 Td[(tji)]TJ /F7 11.955 Tf 11.96 0 Td[(ljj2)]TJ /F7 11.955 Tf 11.95 0 Td[(pj,(3)wherepjisthepricechargedbyserviceproviderj.Theparametertmeasuresthedisutilityofvisitingaserviceproviderwhosehorizontalserviceattributeisnotthe 4Horizontallocationscanbeinterpretedascapturingtastesorphysicalgeographiclocations. 82

PAGE 83

idealattributepreferredbythepatient.Thismodelassumesthatthedisutilitythatpatientssufferfromvisitinganon-ideallocation(i.e.,travelcosts)areconvexinthedistancetraveled.Thisassumptionisrealisticinhealthcaremarkets,inwhichboththecostsoftravelinginemergencysituationsandthecostsofvisitingaproviderwithadifferentspecialtyarelikelyincreasingatanincreasingrateinthedistancetraveled.Additionally,thisassumptionguaranteesequilibriumexistenceifthexedlocationsassumptionofthestylizedmodelwererelaxed.ThreeserviceprovidersarelocatedalongtheHotellingline,withlocationsl1,l2,andl3,wherel1
PAGE 84

locatedatl3.Demandateachlocationisgivenas: D1(p,q)=1(p1,p2,q1,q2)D2(p,q)=2(p2,p3,q2,q3))]TJ /F6 11.955 Tf 11.95 0 Td[(1(p1,p2,q1,q2)D3(p,q)=1)]TJ /F6 11.955 Tf 11.96 0 Td[(2(p2,p3,q2,q3)(3)Itiseasy,yetnontrivial,toshowthatthedemandforaparticularprovider`sserviceisdirectlyaffectonlybyitsownpriceandqualityandthepriceanqualityofitsmoreproximatecompetitors. Figure3-1. PatientChoice Anticipatingtheprot-maximizingqualitychoicesofthethreeindependentserviceproviders,aregulatorsetsserviceprices.Theregulatorsetspricestomaximizetotalsurplus,subjecttoserviceprovidersselectingtheirservicequalitiesindependently. maxp1,p2,p3Z10U(p1,q1)+p1dx+Z21U(p2,q2)+p2dx+Z12U(p3,q3)+p3dx)]TJ /F10 7.97 Tf 18.78 14.94 Td[(3Xqj=1q2j 2subjectto@j @qj=0,j=1,2,3.(3)ThisformulationoftheregulatorsoptimizationproblemspeciesaparticularsequenceofmovesandidentiesaSubgamePerfectNashEquilibriumwithinthatsequentialinteraction.Thetimingisexplicitlygivenas:Stage1)Theregulatorsetstheregulatedprices,Stage2)Eachserviceproviderselectsitsownlevelofquality,independently,Stage3)Consumerspurchaseoneunitofservicefromtheproviderwhoseserviceoffersthegreatestnetutility. 84

PAGE 85

Theanalysispresentedinthissectionconsidersbothstandardizedregulatedpricesthatapplytoallprovidersandindividualizedprices.Whenpricesarestandardized,theregulatorsetsasingleprice,adoptingtheadditionalconstraintp1=p2=p3totheobjectivefunctionpresentedabove.Thesetwouniquesettingsarecomparedanddiscussed.Thesectionthenanalyzestheeffectofamergerbetweenproximatecompetitorsineachsetting.StandardizedPrices:Eachproviderbalancesthecostofqualityimprovementsagainsttheenhancedmarketshareassociatedwiththoseimprovements.Whenthethreeserviceprovidersindependentlysettheirqualitylevels(i.e.,beforeanymergertakesplace),thebenetofqualityimprovementsaredrivenbyown-priceeffects.Theregulatedpriceofaparticularserviceproviderdeterminesthemarginalrevenueofincreasingmarketsharethroughqualityadjustments.Thepre-mergerbestreplyfunctionsaregivenas: q1(p1,p2,p3,q2,q3)=p1 2t(l2)]TJ /F7 11.955 Tf 11.95 0 Td[(l1)q2(p1,p2,p3,q1,q3)=p2 t(l2)]TJ /F7 11.955 Tf 11.96 0 Td[(l1)q3(p1,p2,p3,q1,q2)=p3 2t(l2)]TJ /F7 11.955 Tf 11.95 0 Td[(l1),(3)wherep1=p2=p3.Whentheregulatoradoptsastandardizedregulatedprice.ThesebestreplyfunctionsimplythatserviceprovidersdonotengageinStage3strategicbehavior.Thatis,eachhospitalmakesitsqualitychoiceindependentofthequalitychoicesoftheothertwohospitals.Thesebestreplyfunctions(andtheremainedofthispaper)assumethat,duetosymmetrybetweenHospital1andHospital3,eachislocatedequidistantfromHospital2.Thatis,l2)]TJ /F7 11.955 Tf 12.99 0 Td[(l1=l3)]TJ /F7 11.955 Tf 13 0 Td[(l2.Moreover,duetosymmetricinteractionswithHospital1andHospital3,thelocationofHospital2mustbel2=1 2.Despitelocationsbeingexogenouslydeterminedinthismodel,symmetryofthelocationsofHospital1and 85

PAGE 86

Hospital3wouldarisewithendogenouslocationchoices,providinggeneralizabilityoftheresultsobtained.6Observingtheexogenouslydetermined(symmetric)locationsofthethreeserviceprovidersandanticipatingthebestreplyfunctionofeachrm,theregulatorsetspricestomaximizewelfare,denedasthesumofconsumerandproducersurplus.Whentheregulatorisconstrainedbyarequirementthatp1=p2=p3,theSubgamePerfectNashEquilibriumpriceischaracterizedbythefollowingcondition: @W @p=Z10@q1 @pdx+Z21@q2 @pdx+Z12@q3 @pdx)]TJ /F10 7.97 Tf 18.38 14.94 Td[(3Xj=1qj@qj @p=0.(3)Employingthebestreplyfunctionsidentiedabove,Iamabletoidentifythestandardizedregulatedpriceasp=2(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)2t2 6(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.59 0 Td[(1.Pre-MergerResult1:Theregulatorchoosesastandardizedprice,p=2(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)2t2 6(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.59 0 Td[(1inStage1,andtheserviceprovidersindependentlyselectqualities,(q)(q1,q2,q3)=(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t 6(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.59 0 Td[(1,2(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t 6(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.59 0 Td[(1,(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t 6(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.58 0 Td[(1inStage2.Thisresultsinequilibriummarketshares,(s)(s1,s2,s3)=l2+l1 2)]TJ /F10 7.97 Tf 33.04 4.7 Td[(1 6(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.59 0 Td[(1,l3)]TJ /F9 7.97 Tf 6.59 0 Td[(l1 2+2 6(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.59 0 Td[(1,2)]TJ /F9 7.97 Tf 6.58 0 Td[(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l3 2)]TJ /F10 7.97 Tf 33.03 4.7 Td[(1 6(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.59 0 Td[(1Inequilibrium,allproviderschargethesameprice,asimposedhere,butservicequalityisheterogeneousacrossproviders.Hospital2offerstwicetheservicequalityofitscompetitors.Thisresultisdrivenbyafullcoverageassumption.Hospital2hasmarginalconsumersonbothsides,whereasitscompetitorsonlyhavemarginalconsumerstooneside.Thus,Hospital2experiencestwicethemarginalrevenueofqualityadjustmentsofitscompetitors.Theregulator'sequilibriumpriceisincreasing(decreasing)inboththedistancebetweenHospital1andHospital2andthetransportcostparamtertwheneverthedistancebetweenHospital1andHospital2isgreater(less)than1 3.Increasingthe 6Asmentionedabove,theconvexityofconsumers`travelcostswouldensureequilibriumexistenceinanendogenous-locationssetting. 86

PAGE 87

distancebetweencompetitors`locationsorconsumers`travlecostscanbeinterpretedassofteningcompetition.Thus,whencompetitionissufcientlyweak,increasingitleadstoalowerregulatedprice;whencompetitionissufcientlystrong,increasingitleadstoahigherregulatedprice.Theequilibriumlevelsofservicequalityarealwaysincreasinginthedegreeofcompetition.IndividualizedPrices:Regulatorsmaynotbeconstrainedbyamandatetochargeidenticalpricestoallserviceproviders.Theymaysetunique,individualizedpricesforeachprovider.Whentheregulatorhastheabilitytosetheterogeneousprices,theconstraint,p1,p2,p3,isremovedfromitsoptimizationproblem,implyingthatsocialwelfaremust(weakly)increaseabovethelevelenjoyedwhenpricesarehomogenous.TheStage3bestreplyfunctionsthatcharacterizetheSubgamePerfectNashEquilibriumqualitychoicesaresimilartothosepresentedinthestandardized-pricesetting.Theyonlydifferinoneregard:Changingthepricechargedbyonecompetitoronlyaffectsitsownqualitychoice.Bysymmetry,IonlyneedtoderivetheequilibriumpricesatHospital1andHospital2.Theregulator`soptimalpricesarecharacterizedbythefollowingconditions:Pre-MergerResult2:Theregulatorchoosesaindividualizedprices,(p)(p1,p2,p3)=(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)(l2+l1)t (l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t+1,(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)(l2+l1)t 2[(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t+1],(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)(l2+l1)t (l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t+1inStage1,andtheserviceprovidersindependentlyselectqualities,(q)(q1,q2,q3)=(l2+l1) 2[(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t+1],(l2+l1) 2[(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t+1],(l2+l1) 2[(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t+1]inStage2.Thisresultsinequilibriummarketshares,(s)(s1,s2,s3)=l2+l1 2)]TJ /F10 7.97 Tf 35.55 4.71 Td[(1 4[(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t+1],l3)]TJ /F9 7.97 Tf 6.58 0 Td[(l1 2+1 2[(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t+1],2)]TJ /F9 7.97 Tf 6.59 0 Td[(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l3 2)]TJ /F10 7.97 Tf 35.55 4.71 Td[(1 4[(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t+1]Theequlibriummarketsharesarelikelytobegreaterfortheoutsidecompetitorsandlessfortheinsidecompetitorinthepresenseofindividualized,regulatedprices.Ifthedegreeofhorizontalcompetition,ascapturedbythelocationsofthecompetitors,issufcientlyweak(i.e.,l1<1 6),thisresultisunambiguous.Ifthehorizontallocationsarenotsufcientlyfarapart,however,obtaininganalternativeresultrequiresthatt>l1+4.5 4l1)]TJ /F10 7.97 Tf 6.58 0 Td[(6l21)]TJ /F10 7.97 Tf 6.59 0 Td[(0.5.Theminimumvalueoftthatcansatisfythisconditionis28.9.Thisfeature 87

PAGE 88

isparticularlyinteresting,asgreatervaluesofl1andlowervaluesoftimplydifferentformsofheightenedcompetition.They,however,haveuniqueeffectsontherelationshipbetweenequiliubriummarketsharesunderstnadardizedandindividualized,regulatedprices. Proposition3.1. Ifl1<1 6,theequilibriummarketsharesunderindividualized,regulatedpricesaregreater(less)forHospital1andHospital3(Hospital2).Ifl1>1 6,theequilibriummarketsharesaregreater(less)forHospital1andHospital3(Hospital2)forallvaluesoft<28.9.TheregulatoranticipatestheheightenedmarginalrevenueofqualityimprovementsofHospital2,andsetspricestoequalizequalityacrossproviders.Giventheconvexcostsofqualityimprovements,equalizingservicequalityacrosscompetitorsmaximizessocialwelfare.Consequently,thepriceatHospital2ishalfthepriceatHospitals1and3;patientsatHospital2enjoygreatersurplusthanthosethatpurchaseservicefromoutsideproviders.Theremainderofthischapteranalyzestheeffectsofamergeronregulatedpricesandunregulatedservicequality.Itcomparesthepost-mergerenvironmenttothepre-mergerenvironmentdescribedabove.Onlymergersbetweenmost-proximatecompetitorsareconsidered,asmergersbetweennon-neighboringcompetitorsdonothaveanyeffectonqualitychoicesinthepresenceofregulatedprices.Whenpricesareregulated,regardlessofwhethertheyarehomogenous,mergersbetweennon-neighboringcompetitorshavenoeffect. Proposition3.2. Inthepresenceofregulatedpricing,mergersbetweennon-neighboringcompetitorshavenoeffectofequilibriumqualityprovisionorprices.Post-MergerStandardizedPrices:Asmergersbetweennon-neighboringcompetitorshavenoeffect,andthetwooutsidermsareassumedtobeidentical(exante),itisonlyrelevanttoconsideramergerbetweenHospital1andHospital2.Suchamergerwouldalterthemaximizationproblemofthenewlycombinedentityandalterincentivesfor 88

PAGE 89

qualityadjustments.Thenewlycombinedprovidermaximizesprots: maxq1,q2=p1D1(p,q)+p2D2(p,q))]TJ /F7 11.955 Tf 13.15 8.09 Td[(q21 2)-222()]TJ /F7 11.955 Tf 22.45 8.09 Td[(q22 2.(3)Solvingthemergedprovidersmaximizationproblemidentiesalteredincentivesforqualityprovision.Asdiscussedabove,inthepresenceofregulatedprices,providersmaketheirqualitydecisionsbasedontheirabilitytoattractmarginalconsumers.ThemergedprovidersincentiveregardingmarginalconsumersbetweenHospital1andHospital2areweakened,asanyswitchingonlyaffectsthepricereceived,notactualmarketshare.Whenpricesareequalizedthisaffectdisappears.Post-MergerResult1:Theregulatorchoosesastandardizedprice,p=4(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)2t2 6(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.58 0 Td[(1inStage1,andtheserviceprovidersindependentlyselectqualities,(q)(q1,q2,q3)=0,2(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t 6(l2)]TJ /F9 7.97 Tf 6.58 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.58 0 Td[(1,2(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t 6(l2)]TJ /F9 7.97 Tf 6.59 0 Td[(l1)t)]TJ /F10 7.97 Tf 6.59 0 Td[(1inStage2.Whenstandardizedpricingisadoptedbyorimposedupontheregulator,thereisnoincentivetoprovidequalityatHospital1.Additionally,themarginalrevenueassociatedwithqualityprovisionisidenticalatHospital2andHospital3,aseachisonlyconcernedwithonegroupofmarginalconsumers.Comparingthepre-mergerandpost-mergerprovisionofqualityinaprice-regulatedmarketwithstandardizedpricing,itcanbeseenthatthemergerplacessubstantialupwardpressureonpricesandhasdifferentialeffectsonqualityatthethreeserviceproviders.First,despitethepresenceofaprice-settingregulatorthatseekstomaximizewelfare,themergerstillresultsinupwardpressureonprices.Theregulatordoublesthepriceafterthecombination.Additionally,themergerhaseffectsonqualityprovision.ServicequalityatHospital1fallstozero,atHospital2remainsconstant,andatHospital3doubles.Post-MergerIndividualizedPrices:Aswiththepre-mergerenvironment,itisimportanttoconsidertheeffectofremovingthehomogenouspricesmandateontheregulator.Asbefore,removingtheconstraint,p1=p2=p3,(weakly)improvestotal 89

PAGE 90

surplus.7.Themergerseffectonregulatedpricesandqualityreadjustmentsislessstraightforward.InStage1,theregulatorsetspricestomaximizewelfare,anticipatingsecond-stagebehaviorbytherms.ThemergeraffectstheStage2bestreplyfunctionsofthemerginghospitals,asqualitycompetitionbetweenthemisreducedbythemerger.Itdoesnot,however,affectthebestreplyofthenon-merginghospitalortheoptimalpriceofitsservice.Here,Iexaminetheeffectofthemergeronequilibriumregulatedpricesatthemerginghospitals.Holdingpricesattheirpre-mergerlevels,themarginalwelfareeffectofincreasingthepricesatHospital1andHospital2canbeidentiedas: @W @p1=Z21@q2 @p1dx+[p1)]TJ /F7 11.955 Tf 11.96 0 Td[(p2]@1 @q2@q2 @p1+[p2)]TJ /F7 11.955 Tf 11.96 0 Td[(p3]@2 @q2@q2 @p1,and(3) @W @p2=Z10@q1 @p2dx+[p1)]TJ /F7 11.955 Tf 11.95 0 Td[(p2]@1 @q1@q1 @p2.(3)Itissimpletoobservethatthemerger'seffectontheoptimalregulatedpriceatHospital2isnegative;theregulatorwillreducethepriceatHospital2afterthetransaction.Themerger'seffectonthepriceatHospital1is,however,ambiguous.Whilethersttermof@W @p1,asevaluatedatthepre-mergerprices,isnegative,thesecondtwotermsarepositive.Balancingtheseeffectsrequiresprecisespecicationofthehorizontallocationsandthetransportcostparametert.Post-MergerResult2:ThemergerhasanambiguouseffectontheregulatedpriceatHospital1,anegativeeffectontheregulatedpriceatHospital2,andnoeffectontheregulatedpriceatHospital3.Withoutassigningnumericalvaluestotheexogenousparametersofthemodel,theeffectofthemergeronservicequalitiesisambiguous. 7Inthepre-mergerenvironment,priceschangedwhenthisconstraintwasremoved,implyingthatwelfarestrictlyimproved 90

PAGE 91

3.4ConclusionTheFederalTradeCommissionreviewshospitalmergersandchoosestoblockmergersthatarelikelytorestrictcompetition.TheanalysisthattheFTCconductsislaidoutinthe2010HorizontalMergerGuidelines.Theseguidelinesfocusonthelikelyeffectofaproposedmergeronprices;mergersthatareexpectedtoincreasepricesthroughtheeliminationofaviablecompetitorareoftenblocked.Theguidelinesmentionproductdifferentiationasafactorthatsoftenscompetitionandmakespost-mergerpriceincreaseslesslikely.Theydonot,however,addressthelikelyeffectsofatransactiononpost-mergerproductdifferentiation,whichisasubstantialdeterminantofpatientswelfare.Theanalysispresentedinthispaperisintendedtoinformpractitionersinthisarea.Thispaperexaminestheeffectofmergersonservicequalityinprice-regulatedmarkets.Iadoptathree-stagegame,inwhichtimingisasfollows:State1)Theregulatorsetstheregulatedprices,Stage2)Eachserviceproviderselectsitsownlevelofquality,indepently,Stage3)Consumerspurchaseoneunitofservicefromtheproviderwhoseserviceoffersthegreatestnetutility.Inthissetting,theeffectofamergerbetweenneighboringcompetitorsdependsonthenatureofpriceregulation.Whenregulatedpricesmustbeequalacrossallproviders,themergerresultsinhigherpricesatallserviceproviders,butithasdifferentialeffectsonservicequality.Thequalityatoneofthemergingprovidersfalls,andattheother,remainsconstant.Interestingly,theservicequalityatthenon-mergingproviderincreases.Whenregulatedpricesmayvaryacrossproviders,theeffectofthemergerisambiguous.Theregulatedpriceandservicequalityatthenon-mergingproviderremainconstant.Theregulatedpriceshowevermayincreaseordecrease,leadingtoambiguouseffectsonqualitychoicesinStage2.Whenaregulatormaychoosetosetstandardizedorindividualizedprices,itisalways(weakly)welfareenhancingtoselectthelatter.Standardizedpricesimposean 91

PAGE 92

additionalconstraintontheregulatorswelfareoptimizationproblem,andthus,cannotenhancewelfare. 92

PAGE 93

APPENDIX:MATHEMATICALDERIVATIONS Corollary1(i):When^eB>eG,andtheentrantisabletoprotablyenter,anyin-creaseinthemarginalcostofgreenwashreducestheamountofgreenwashundertakenbytheincumbentatadecreasingrate.TheFirstOrderCondition(FOC)fortheprot-maximizinglevelofreportedharmisgivenby: V2(4^eB)]TJ /F10 7.97 Tf 6.59 0 Td[(7eG) (4^eB)]TJ /F9 7.97 Tf 6.59 0 Td[(eG)3=k(e)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB).TotallydifferentiatingtheFOCw.r.t.themarginalcostofgreenwashingprovides: V2(80eG)]TJ /F10 7.97 Tf 6.58 0 Td[(32^eB) (4^eB)]TJ /F9 7.97 Tf 6.58 0 Td[(eG)4@^eB @k=(e)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB))]TJ /F7 11.955 Tf 11.96 0 Td[(k@^eB @k.Ask>@2TR @^eB2inordertosatisfytheSecondOrderSondition(SOC),@^eB @k>0.TakingthesecondordertotalderivativeoftheFOCallowsustoverifytheconcavityoftheoptimal^eB(k)function.Ageneralrepresentationofthersttotalderivativeisgivenas: )]TJ /F15 7.97 Tf 10.49 4.71 Td[(@2TR @^eB2)]TJ /F15 7.97 Tf 6.68 -4.58 Td[(@^eB @k2=(eB)]TJ /F8 11.955 Tf 15.08 0 Td[(^eB))]TJ /F7 11.955 Tf 11.95 0 Td[(k@^eB @k.Thus,thesecondtotalderivativew.r.t.kis: )]TJ /F15 7.97 Tf 10.5 4.7 Td[(@3TR @^eB3)]TJ /F15 7.97 Tf 6.67 -4.57 Td[(@^eB @k2)]TJ /F15 7.97 Tf 13.15 4.7 Td[(@2TR @^eB2@2^eB @k2=)]TJ /F8 11.955 Tf 9.3 0 Td[(2@^eB @k)]TJ /F7 11.955 Tf 11.95 0 Td[(k@2^eB @k2. k)]TJ /F15 7.97 Tf 13.15 4.71 Td[(@2TR @^eB2@2^eB @k2=@3TR @^eB3)]TJ /F15 7.97 Tf 6.68 -4.57 Td[(@^eB @k2)]TJ /F8 11.955 Tf 11.95 0 Td[(2@^eB @k.Thus,weobservethat@2^eB @k2<0.Asaconsequence,withinCase1,increasesinthemarginalcostofgreenwashingincreasetheincumbent'sprot-maximizingreportatadecreasingrate.Corollary1(ii):When^eB>eG,andtheentrantisabletoprotablyenter,anyincreaseinthereportedlevelofenvironmentalharmbytheincumbentdecreasesaggregateenvironmentalharmatadecreasingrate.Thatis,greenwashbytheincum-bentincreasesaggregateharmatanincreasingrate.Additionally,increasesintheincumbent'smarginalcostofgreenwashingreduceaggregateharmatadecreasingrate. 93

PAGE 94

TakingtherstorderpartialderivativeoftheaggrehateenvironmentalharminCase1,Iobtain: @H @^eB=)]TJ /F10 7.97 Tf 6.58 0 Td[(2V(2eB+eB) (4^eB)]TJ /F9 7.97 Tf 6.59 0 Td[(eG)2<0.TakingthesecondorderpartialderivativeoftheaggrehateenvironmentalharminCase1,Iobtain: @2H @^eB2=16V(2eB+eB) (4^eB)]TJ /F9 7.97 Tf 6.59 0 Td[(eG)3>0.WithinCase1,increasesintheincumbent'sreportdecreaseharmatadecreasingrate,oralternatively,increasesintheprot-maximizinglevelofgreenwashincreaseaggregateenvironmentalharmatanincreasingrate.Ageneralizedexpressionforaggregateenvironmentalasafunctionofthemarginalcostofgreenwashingisgivenas,H=H(^eB(k)).Differentiatingthisexpressiongives: @H @k=@H @^eB@^eB @k.As@H @^eB<0,and@^eB @k>0,weobservethat@H @k<0.Takingthesecondderivativeofthegeneralexpressionforharmgives: @2H @k2=@H @^eB@2^eB @k2+@2H @^eB2@^eB @k>0.Thus,withinCase1,increasesintheincumbent'smarginalcostofgreenwashdecreaseharmatadecreasingrate.Corollary1(iii):When^eB>eG,andtheentrantisabletoprotablyenter,thereportedlevelofharmbytheincumbentthatmaximizesconsumersurplusislessthaneB,and(a)when^eB=4eB)]TJ /F10 7.97 Tf 6.59 0 Td[(3eG 16resultsinentrybythegreenrm,^eB=4eB)]TJ /F10 7.97 Tf 6.59 0 Td[(3eG 16maximizesconsumersurplus,(b)when^eB=4eB)]TJ /F10 7.97 Tf 6.58 0 Td[(3eG 16doesnotresultinentrybythegreenrm,areportedlevelofharmjustabovetheentrythresholdlevelofharmmaximizesconsumersurplus.DeningconsumersurplusasCSR^0V)]TJ /F6 11.955 Tf 12.28 0 Td[(eB)]TJ /F7 11.955 Tf 12.27 0 Td[(pBd+R~^V)]TJ /F6 11.955 Tf 12.28 0 Td[(eG)]TJ /F7 11.955 Tf 12.27 0 Td[(pGd,itfollowsthat:@CS @^eB=R^0)]TJ /F15 7.97 Tf 10.49 5.11 Td[(@pB @^eBd+(V)]TJ /F8 11.955 Tf 10.93 2.66 Td[(^eB)]TJ /F7 11.955 Tf 10.2 0 Td[(pB)@^ @^eB+[^eB)]TJ /F8 11.955 Tf 13.34 0 Td[(^eB]@^ @^eB^+R~^)]TJ /F15 7.97 Tf 10.5 5.11 Td[(@pG @^eBd+(V)]TJ /F8 11.955 Tf 10.93 2.66 Td[(~eG)]TJ /F7 11.955 Tf 10.2 0 Td[(pG)@~ @^eB 94

PAGE 95

-(V-eG)]TJ /F7 11.955 Tf 11.95 0 Td[(pG)@^ @^eB.AsV)]TJ /F8 11.955 Tf 12.28 2.66 Td[(^^eB)]TJ /F7 11.955 Tf 11.56 0 Td[(pB=V)]TJ /F8 11.955 Tf 12.29 2.66 Td[(^eG)]TJ /F7 11.955 Tf 11.56 0 Td[(pG,andV)]TJ /F8 11.955 Tf 12.29 2.66 Td[(~eG)]TJ /F7 11.955 Tf 11.56 0 Td[(pG=0,themarginaleffectof^eBonCSreducesto: @CS @^eB=R^0)]TJ /F15 7.97 Tf 10.49 5.12 Td[(@pB @^eBd+[^eB)]TJ /F7 11.955 Tf 11.96 0 Td[(eB]@^ @^eB^+R~^)]TJ /F15 7.97 Tf 10.49 5.12 Td[(@pG @^eBd @CS @^eB=V2(4eB)]TJ /F10 7.97 Tf 6.58 0 Td[(16^eB)]TJ /F10 7.97 Tf 6.59 0 Td[(3eG) (4^eB)]TJ /F9 7.97 Tf 6.58 0 Td[(eG)3Asthesecondpartialisclearlynegative,theconsumersurplusfunction(inCase1)obtainsamaximumat^eB=4eB)]TJ /F10 7.97 Tf 6.58 0 Td[(3eG 16.Itispossible,however,thatthismaximumisnotacheivable(withinCase1)ifeBandeGareparticularlyclosetogetherofthexedcostsofentryaresufcientlyhigh.Corollary2:If^eB>eG,whengreenwashissufcientlystrongtoforeclosethemarkettoagreenerentrant,theincumbent'sprotsexhibitanupwardjump,andthemarginalbenetofgreenwashexhibitsanupwardjump.Together,thesecharacteristicsimplythattheprot-maximizingreportjumpsdown.DeningTRiasthetotalrevenuethattheincumbentformreceivesinCasei,itcanbeshownthattotalrevenuejumpsupwhengrenwashissufcientlystrongtoshiftthemarketfromCase1toCase2bydeterringentry: TR10.Additionally,itcanbeshownthatthemarginalbenetofgreenwashisgreateroncetheincumbentisabletosecuremonopolystatus.DeningMBiasthemarginalbenetofgreenwashwithinCasei,weobserve: MB1
PAGE 96

16^eB3)]TJ /F8 11.955 Tf 11.96 0 Td[(28^eB2eG<64^eB3)]TJ /F8 11.955 Tf 11.95 0 Td[(48^eB2eG+12^eBe2G)]TJ /F7 11.955 Tf 11.96 0 Td[(e2G )]TJ /F8 11.955 Tf 9.29 0 Td[(48^eB3+20^eB2eG)]TJ /F8 11.955 Tf 11.95 0 Td[(12^eBe2G+e2G<0Thus,whengreenwashissufcientlystrongtodeterentry(betweenCases1and2),theincumbent'sprotsjumpup,asdoesthemarginalbenetitreceivesfromgreenwashing.Corollary3(i):Whentheentrantisnotabletoprotablyenter,anyincreaseinthemarginalcostofgreenwashreducestheamountofgreenwashundertakenbytheincumbentatadecreasingrate.TheFirstOrderCondition(FOC)fortheprot-maximizinglevelofreportedharmisgivenby: V2 4^eB2=k(eB)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB).TotallydifferentiatingtheFOCw.r.t.themarginalcostofgreenwashingprovides: )]TJ /F9 7.97 Tf 6.58 0 Td[(V2 2^eB3=(eB)]TJ /F8 11.955 Tf 15.09 0 Td[(^eB))]TJ /F7 11.955 Tf 11.95 0 Td[(k@^eB @k k)]TJ /F9 7.97 Tf 16.49 4.71 Td[(V2 2^eB3@^eB @k=(eB)]TJ /F8 11.955 Tf 15.08 0 Td[(^eB)Ask>@2TR @^eB2inordertosatisfytheSecondOrderSondition(SOC),@^eB @k>0.TakingthesecondordertotalderivativeoftheFOCallowsustoverifytheconcavityoftheoptimal^eB(k)function: 1+3V2 2^eB4@^eB @k+k)]TJ /F9 7.97 Tf 16.49 4.71 Td[(V2 2^eB3@2^eB @k2=)]TJ /F15 7.97 Tf 10.49 5.11 Td[(@^eB @k,Thus,weobservethat@2^eB @k2<0.Asaconsequence,withinCases2and3,increasesinthemarginalcostofgreenwashingincreasetheincumbent'sprot-maximizingreportatadecreasingrate.Corollary3(ii):Withintherangeofmarginalcostsofgreenwashthatgivesrisetoamonopolymarketstructure,increasesinthemarginalcostofgreenwashdecreasesaggregateenvironmentalharmatadecreasingrate.FromthemethematicalderivationforCorollary2: @H @k=@H @^eB@^eB @k.As@H @^eB<0,and@^eB @k>0,weobservethat@H @k<0.Takingthesecondderivativeofthegeneralexpressionforharmgives: 96

PAGE 97

@2H @k2=@H @^eB@2^eB @k2+@2H @^eB2@^eB @k>0.Thus,withinCases2and3,increasesintheincumbent'smarginalcostofgreenwashdecreaseharmatadecreasingrate.Corollary3(iii):Consumersurplusjumpsdownwheneverentryisforeclosed,regardlessofwhether^eB>eG.Additionally,withintherangeofmarginalcostsofgreenwashthatgivesrisetoamonopolymarketstructure,increasesinthemarginalcostofgreenwashincreaseconsumersurplusatadecreasingrate.DeningconsumersurplusinthemonopolycasesasCSR^0V)]TJ /F6 11.955 Tf 11.67 0 Td[(eB)]TJ /F7 11.955 Tf 11.67 0 Td[(pBd,anddeningconsumersurplusinCase4asCSR^0V)]TJ /F6 11.955 Tf 11.63 0 Td[(eG)]TJ /F7 11.955 Tf 11.64 0 Td[(pGd+R~^V)]TJ /F6 11.955 Tf 11.64 0 Td[(eB)]TJ /F7 11.955 Tf 11.64 0 Td[(pBd,itisclearthatconsumersurplusjumpsdownanytimethatentryisforeclosed.First,IdemonstratethatConsumersurplusjumpsdownwhenthemarketmovesfromCase1toCase2.Here,Idene^1and~1asintheCorollary3proofand^2isthelocationoftheconsumerthatisindifferentbetweenconsumingfromamonopolistandabstainingfromconsumption.~1>^2>^1.Forsimplicity,themonopolypriceisdenotedaspM.Thelosstomonopoly(betweencase1andCase2)foragiven^eBisgivenas: R^10pM)]TJ /F7 11.955 Tf 11.95 0 Td[(pBd+R^2^1(eB)]TJ /F7 11.955 Tf 11.96 0 Td[(eG)+(pM)]TJ /F7 11.955 Tf 11.96 0 Td[(pG)d+R~1^2V)]TJ /F6 11.955 Tf 11.96 0 Td[(eG)]TJ /F7 11.955 Tf 11.96 0 Td[(pGd>0ConsumersurplusjumpsdownwhenthemarketarrangementshiftsfromthatdescribedbyCase1tothatdescribedbyCase2.Next,IdemonstratethatConsumersurplusjumpsdownwhenthemarketmovesfromCase3toCase4.Here,Idene^4and~4asthelocationoftheconsumerthatisindifferentbetweenthegreenlowqualityvariantandtheincumbent'svariantandthelocationoftheconsumerthatisindifferentbetweenconsumingfromtheincumbentandabstainingfromconsumption,respectively.Also,^3isthelocationoftheconsumerthatisindifferentbetweenconsumingfromamonopolistandabstainingfromconsumption.~4>^3>^4.Forsimplicity,themonopolypriceisdenotedaspM.Thelosstomonopoly(betweencase1andCase2)foragiven^eBisgivenas: R^40(eB)]TJ /F7 11.955 Tf 11.95 0 Td[(eG)+(pM)]TJ /F7 11.955 Tf 11.95 0 Td[(pG)d+R^3^4pM)]TJ /F7 11.955 Tf 11.96 0 Td[(pBd+R~4^3V)]TJ /F6 11.955 Tf 11.96 0 Td[(eB)]TJ /F7 11.955 Tf 11.96 0 Td[(pBd>0 97

PAGE 98

ConsumersurplusjumpsdownwhenthemarketarrangementshiftsfromthatdescribedbyCase4tothatdescribedbyCase3.Thus,wheneverthemarketisforeclosed,consumersurplusjumpsdown.Finally,withinthemonopolymarket,consumersurplusincreasesatadecreasingrateas^eBincreases: @CS @^eB=V2(eB)]TJ /F10 7.97 Tf 9.01 0 Td[(^eB) 4^eB3>0 @2CS @^eB2=V2(2^eB)]TJ /F10 7.97 Tf 6.58 0 Td[(3eB) 4^eB4<0Corollary4(i):When^eBTR4a V2 4^eB>4V2(eG)]TJ /F10 7.97 Tf 9.01 0 Td[(^eB)eG (4eG)]TJ /F10 7.97 Tf 9.01 0 Td[(^eB)2^eB 16e2G)]TJ /F8 11.955 Tf 11.95 0 Td[(8^eBeG+^eB2>16e2G)]TJ /F8 11.955 Tf 11.96 0 Td[(16^eBeG 8^eBeG+^eB2>0Additionally,itcanbeshownthatthemarginalbenetofgreenwashisgreateroncetheentrantisabletoenter.DeningMBiasthemarginalbenetofgreenwashwithinCasei,weobserve: MB3)]TJ /F8 11.955 Tf 12.43 0 Td[(^eB3+12^eB2eG)]TJ /F8 11.955 Tf 11.96 0 Td[(48^eBe2G+64e3G 20^eB2eG+^eB3>0 98

PAGE 99

Thus,whengreenwashissufcientlystrongtofacilitateentry(betweenCases3and4),theincumbent'sprotsjumpdown,butthemarginalbenetitreceivesfromgreenwashingjumpsup.WecanobservethatwithinSubcase4a,increasesinthemarginalcostofgreenwashincreasetheincumbent'sprot-maximizinglevelofreportedharmatadecreasingrate.Totallydifferentiatingtheincumbent'sFOCforprotmaximizationinStage1,Iobtain: )]TJ /F10 7.97 Tf 6.59 0 Td[(8V2(6^eB2eG)]TJ /F10 7.97 Tf 6.59 0 Td[(3^eB3+16e3G)]TJ /F10 7.97 Tf 6.59 0 Td[(16^eBe2G) (4eG)]TJ /F10 7.97 Tf 9.01 0 Td[(^eB)4^eB3@^eB @k=(e)]TJ /F8 11.955 Tf 15.08 0 Td[(^eB))]TJ /F7 11.955 Tf 11.95 0 Td[(k@^eB @k.Ask>@2TR @^eB2inordertosatisfytheSecondOrderSondition(SOC),@^eB @k>0.TakingthesecondordertotalderivativeoftheFOCallowsustoverifytheconcavityoftheoptimal^eB(k)function.Ageneralrepresentationofthersttotalderivativeisgivenas: )]TJ /F15 7.97 Tf 10.49 4.7 Td[(@2TR @^eB2)]TJ /F15 7.97 Tf 6.68 -4.57 Td[(@^eB @k2=(eB)]TJ /F8 11.955 Tf 15.08 0 Td[(^eB))]TJ /F7 11.955 Tf 11.95 0 Td[(k@^eB @k.Thus,thesecondtotalderivativew.r.t.kis: )]TJ /F15 7.97 Tf 10.5 4.71 Td[(@3TR @^eB3)]TJ /F15 7.97 Tf 6.67 -4.57 Td[(@^eB @k2)]TJ /F15 7.97 Tf 13.15 4.71 Td[(@2TR @^eB2@2^eB @k2=)]TJ /F8 11.955 Tf 9.3 0 Td[(2@^eB @k)]TJ /F7 11.955 Tf 11.95 0 Td[(k@2^eB @k2. k)]TJ /F15 7.97 Tf 13.15 4.71 Td[(@2TR @^eB2@2^eB @k2=@3TR @^eB3)]TJ /F15 7.97 Tf 6.68 -4.58 Td[(@^eB @k2)]TJ /F8 11.955 Tf 11.95 0 Td[(2@^eB @k.Thus,weobservethat@2^eB @k2<0.Asaconsequence,withinSubcase4a,increasesinthemarginalcostofgreenwashingincreasetheincumbent'sprot-maximizingreportatadecreasingrate.Corollary4(ii):When^eB0,weobservethat@H @k<0.Takingthesecondderivativeofthegeneralexpressionforharmgives: @2H @k2=@H @^eB@2^eB @k2+@2H @^eB2@^eB @k>0. 99

PAGE 100

Thus,withinSubcase4a,increasesintheincumbent'smarginalcostofgreenwashdecreaseharmatadecreasingrate.Corollary4(iii):When^eB0Anyincreaseintheincumbent'sreportedlevelofharm(inSubcase4a)resultsleadstoimprovedconsumerwelfare.Corollary5:When^eB0 100

PAGE 101

REFERENCES Adams,WilliamJames,andJanetYellen.1976.Commoditybundlingandtheburdenofmonopoly.TheQuarterlyJournalofEconomics90(3):475. Anderson,Simon,Foros,andHansJarleKind.2010.Hotellingcompetitionwithmulti-purchasing:timemagazine,newsweek,orboth? Aoki,Reiko,andThomasPrusa.1996.Sequentialversussimultaneouschoicewithendogenousquality.InternationalJournalofIndustrialOrganization15(1):103. Arora,Seema,andShubhashisGangopadhyay.1995.Towardatheoreticalmodelofvoluntaryovercompliance.JournalofEconomicBehaviorandOrganization28(3):289. Bradlow,Eric,andVithalaRao.2000.AhierarchicalBayesmodelforassortmentchoice.JournalofMarketingResearch37(2):259. Brekke,Kurt,SicilianiLuigi,andOddRuneStraume.2008.Competitionandwaitingtimesinhospitalmarkets.JournalofPublicEconomics92(7):1607. Brekke,Kurt,SicilianiLuigi,andOddRuneStraume.2011.HospitalCompetitionandQualitywithRegulatedPrices.TheScandinavianJournalofEconomics113(2):444. Carlton,DennisW.2010.Revisingthehorizontalmergerguidelines.JournalofCompetitionLawandEconomics6(3):619. Chung,Jaihak,andVithalaRao.2003.Ageneralchoicemodelforbundleswithmultiple-categoryproducts:Applicationtomarketsegmentationandoptimalpricingforbundles.JournalofMarketingResearch40(2):115. Coate,Malcolm,andShawnUlrick.2006.TransparencyattheFederalTradeCommission:Thehorizontalmergerreviewprocess1996-2003.AntitrustLawJournal73(2):531. Conrad,Klaus.2005.Pricecompetitionandproductdifferentiationwhenconsumerscarefortheenvironment.EnvironmentalandResourceEconomics31(1):1. Corts,Kenneth.2011.Optimal(andnite)penaltiesforfalseadvertising. Courty,Pascal.2003.Someeconomicsofticketresale.TheJournalofEconomicPerspectives17(2):85. Cremer,Helmuth,andJacques-FranThisse.1999.Onthetaxationofpollutingproductsinadifferentiatedindustry.EuropeanEconomicReview43:575. Cremer,MarchandMaurice,Helmuth,andJacques-FranThisse.1991.Mixedoligopolywithproductdifferentiation.InternationalJournalofIndustrialOrganization9(1):43. 101

PAGE 102

Dixit,Avinash,andVictorNorman.1978.Advertisingandwelfare.TheBellJournalofEconomics9(1):1. Eriksson,Clas.2004.Cangreenconsumerismreplaceenvironmentalregulation?-Adifferentiatedproductsexample.ResourceandEnergyEconomics26(3):281. Gabszewicz,Jaskold,andJacques-FranThisse.1979.Pricecompetition,quality,andincomedisparities.JournalofEconomicTheory20(3):340. Glaeser,Edward,andGergelyUjhelyi.2010.Towardatheoreticalmodelofvoluntaryovercompliance.JournalofPublicEconomics94(3):247. Guiltinan,Joseph.1987.Thepricebundlingofservices:anormativeframework.TheJournalofMarketing51(2):74. Hattori,Keisuke,andKeissakuHigashida.2011.Misleadingadvertisingandminimumqualitystandards. Hattori,Keisuke,andKeissakuHigashida.2012.Misleadingadvertisinginduopoly.CanadianJournalofEconomics45(3):1154. Heijen,Pim.2007.Informativeadvertisingbyanenvironmentalgroup. Heijen,Pim,andLambertSchoonbeek.2008.Environmentalgroupsinmonopolisticmarkets.EnvironmentalandResourceEconomics39(4):379. Ho,Vivian,andBartonHamilton.2000.Hospitalmergersandacquisitions:Doesmarketconsolidationharmpatients?JournalofHealthEconomics19(5):767. K,Bagwell.2007.Theeconomicanalysisofadvertising.InHandbookofIndustrialOrganization,editedbyMarkArmstrongandRobertPorter,1701.Amsterdam,Noord-Holland:Elsevier. Kessler,Daniel,andJeffreyGeppert.2005.Theeffectsofcompetitiononvariationinthequalityandcostofmedicalcare.JournalofEconomics&ManagementStrategy14(3):575. Kessler,Daniel,andMarkMcClellan.2000.Ishospitalcompetitionsociallywasteful?TheQuarterlyJournalofEconomics115(2):577. Kotowitz,Yehuda,andFrankMathewson.1979.Advertising,consumerinformation,andproductquality.TheBellJournalofEconomics10(2):284. Krueger,Alan.2005.Theeconomicsofrealsuperstars:Themarketforrockconcertsinthematerialworld.JournalofLaborEconomics23(1):1. Lewbel,Arthur.1985.Bundlingofsubstitutesorcomplements.InternationalJournalofIndustrialOrganization3(1):101. 102

PAGE 103

Lyon,Thomas,andJohnMaxwell.2011.Greenwash:corporateenvironmentaldisclosureunderthreatofaudit.JournalofEconomicsandManagementStrategy20(1):3. Mankiw,Gregory,andMichaelWhinston.1986.Freeentryandsocialinefciency.TheRANDJournalofEconomics17(1):48. Mason,Charles.2011.Eco-labelingandmarketequilibriawithnoisycerticationtests.EnvironmentalandResourceEconomics48(4):537. Moraga-Gonzalez,JoseLuis,andNoemiPadron-Fumero.2002.Environmentalpolicyinagreenmarket.EnvironmentalandResourceEconomics22(3):419. Mukamel,Dana,JackZwanziger,andAnilBamezai.2002.Hospitalcompetition,resourceallocationandqualityofcare.BMCHealthServicesResearch2(1):1. Norman,George,andLynnePepall.2000.ProtablemergersinaCournotmodelofspatialcompetition.SouthernEconomicJournal66(3):667. Salop,Steven.1979.Strategicentrydeterrence.AmericanEconomicReview69(2):335. Sari,Nazmi.2002.Docompetitionandmanagedcareimprovequality?Healtheconomics11(17):571. Schmalensee,Richard.1982.Advertisingandentrydeterrence:Anexplanatorymodel. Schmalensee,Richard.1984.Gaussiandemandandcommoditybundling.TheJournalofBusiness57(1):s211S230. Tay,Abigail.2003.Assessingcompetitioninhospitalcaremarkets:theimportanceofaccountingforqualitydifferentiation.TheRANDJournalofEconomics34(4):786. vanderMade,Allard,andLambertSchoonbeek.2009.Entryfacilitationbyenvironmentalgroups.EnvironmentalandResourceEconomics43(4):457. Venkatesh,R.,andRabikarChatterjee.2006.Bundling,unbundling,andpricingofmultiformproducts:Thecaseofmagazinecontent.JournalofInteractiveMarketing20(2):21. Venkatesh,R.,andWagnerKamakura.2003.OptimalBundlingandPricingunderaMonopoly:ContrastingComplementsandSubstitutesfromIndependentlyValuedProducts.TheJournalofBusiness76(2):211. Volpp,Kevin,JonathanKetcham,AndrewEpstein,andSankeyWilliams.Theeffectsofpricecompetitionandreducedsubsidiesforuncompensatedcareonhospitalmortality.HealthServicesResearch40(4):1056. 103

PAGE 104

Wauthy,Xavier.1996.Qualitychoiceinmodelsofverticaldifferentiation.TheJournalofIndustrialEconomics44(3):345. 104

PAGE 105

BIOGRAPHICALSKETCH ThomasKnightearnedhisBachelorofArtsdegreeineconomicsandenvironmentalstudiesatNewCollegeofFloridain2007.HereceivedhisMasterofSciencedegreeineconomics:competitionandregulationfromTilburgUniversityin2009.In2009,heenteredthedoctoralprogramineconomicsattheUniversityofFlorida,wherehespecializedingametheory,industrialorganization,andexperimentaleconomics.Dr.Knightreceivednumerousawardswhilepursuinghisdoctorate,includingauniversity-wideGraduateTeachingAwardin2013.Additionally,hereceivedresearchandtravelsupportfromthePatriciaL.PaceyFellowship,theMadelynM.LockhartEndowment,andtheUniversityofFloridaGraduateStudentCouncil.Withthisnancialsupport,hewasabletopublishresearchinManagerialandDecisionEconomicsandtheTexasIntellectualPropertyLawJournal.Healsopresentedhisresearchatanumberofeconomicconferences,includingtheSouthernEconomicAssociationConference,PublicChoiceSocietyAnnualMeeting,MidwestEconomicsTheoryMeeting,andAssociationofPublicEconomicTheoryConference.Whilepursuinghisdegree,Dr.KnightservedasanInstructorofEconomicsattheUniversityofFlorida,anAdjunctInstructorofEconomicsatNewCollegeofFlorida,andaSummerResearchAssistantattheUnitedStatesFederalTradeCommission.Dr.Knight'sdissertation,ThreeEssaysonProductDifferentiationandStrategy,wassupervisedbyDr.JonathanHamilton.Upongraduation,Dr.KnightwilljointheeconomicsfacultyattheUniversityofFlorida. 105