产业 调 查 报 告 现状

营销与策划一班 文化

郭颖

一．文化产业的现状

二十一世纪的今天，当今世界经济结构发生了变化，文化产业在经济上非常重要，文化上的强国将成为经济强国。如果一个国家不能创造出自己的文化内容而使本国的内容资源不足的话，将遭遇严重的文化独立性危机。上个世纪七八十年代，世界经济以制造业为中心；上个世纪九十年代，世界经济以服务业和知识为基础；二十一世纪则是以知识产权为基础的内容产业经济。文化产业是投入少、产品附加值高的产业。文化产业范围内的各产业相互联结，其“波及效果”会带动其他类别的产业。

我们国家和社会对文化产业的正确认识对文化产业的发展也非常重要。党的十六大报告中提出，“发展文化产业是市场经济条件下繁荣社会主义文化、满足人民群众精神文化需求的重要途径”，“完善文化产业政策，支持文化产业发展，增强中国文化产业的整体实力和竞争力”。十六届五中全会也指出：“要深化文化体制改革，积极发展文化事业和文化产业，创造更多更好适应人民群众需求的优秀文化产品”.

文化产业是非常独特而又极其复杂的新兴产业,是文化发展到现代历史阶段的产物，但并非所有文化都可以发展为产业。文化部文化产业司司长王永章说，文化产业是指从事文化生产和提供文化服务的经营性行业，文化产业是和文化事业相对应的概念，都是社会文化建设的重要组成部分，文化产业是社会生产力发展的必然要求，是随着社会主义市场经济体制的逐步完善和现代生产方式的不断进步而发展起来的新兴产业。这是今年9月文化部对文化产业的最新界定。文化产业分为影视业、音像业、文化娱乐业、文化旅游业、网络文化业、图书报刊业、文物和艺术品业、艺术培训业等9大门类。而实际上文化产业的范围远远不止以上门类，我国文化产业的范围最起码还应该包括新闻出版、广播电视、文学艺术、信息产业的一部分。

二．发展文化产业的重大意义

党的十七届五中全会提出：“推动文化产业成为国民经济支柱性产业。”当今世界正处在大发展大变革大调整时期，文化与经济、政治等相互交融，日益成为经济社会发展的重要战略资源；国家之间综合国力的激烈竞争，正日益聚集于以文化为核心的软实力的竞争。资源消耗低、环境污染小、以创业为核心的文化产业，正在成为转变经济发展方式的新亮点。

文化产业既凝聚人心，又关切民生，还可以直接贡献经济增长。文化表面看是软实力，实际却是硬实力。国际经验表明，越是发达

国家，文化消费比重越高，文化产业对GDP的贡献越大。据研究，当人均GDP突破3000美元以后，社会对文化产品的消费会有大的突破，人均GDP越高，文化消费占的比例越大。

近年来，我国文化产业的规模不断扩大，文化产业增加值占GDP的比重稳步提升，经济效益大幅提高，文化体制改革成效显著，娱乐文化服务、会议及展览服务等新兴文化产业发展迅速。但我国文化产业发展还存在一些问题，如：我国文化产业的发展水平还不高、活力还不强，与人民群众日益增长的精神文化需求还不相适应，与日趋完善的社会主义市场经济体制还不相适应，与现代科学技术迅猛发展及广泛应用还不相适应，与我国对外开放不断扩大的新形势还不相适应。因此，应大力推动文化产业发展，使文化产业成为经济发展的新引擎。

三．当前我国文化产业发展的问题

1、统一、规范、竞争、有序的现代文化市场体系尚未建立，制约了文化产业的正常发展。 成熟的现代文化市场体系包括文化产品市场、文化服务市场、文化要素市场。建立并完善我国的文化市场体系对文化产业的发展具有深远的意义。文化市场是文化产业不可或缺的重要组成部分，健全的文化市场能够促进各类文化产品和市场要素的自由流动，实现文化资源的优化配置，拓展文化产业的发展空间。

当前我国文化市场的发育还不够完善，文化产品市场、文化服务市场不够发达，文化要素市场的发展就更为滞后，例如，资金市场、设施市场、人才劳务市场、中介市场、产权交易市场等急需建设和发展。这种状况无疑阻碍了文化产品和服务的生产和流通，限制了我国文化市场、文化产业的深入发展。

2、丰富的文化资源未得到有效利用。 我国有极为丰富的文化资源，历史文化资源方面，重要历史人物、重大历史事件不胜枚举；民族文化资源方面，56个民族的民族文化资源多姿多彩，极为丰富；现代文化资源方面，在20世纪，中国在政治上经历了帝制、总统制、共和制，在经济上发生了从小农经济、计划经济到市场经济的巨大转型，在文化上则是各种社会思潮、思想流派异彩纷呈。但是这些丰富的文化资源很少得到有效的开发和利用。流传久远的《花木兰》故事，让许多人熟视无睹，但在被好莱坞加工成动画片后，在世界范围内取得了票房丰收。中华民族五千年来形成的深厚文化积累、多样文化形态，是怎么估计也不过分的宝贵文化资源。如何通过发展文化产业等途径，把我国令人称羡的文化资源优势转变为文化产业优势，是一个亟待解决的现实课题。

3、文化产业法规建设滞后。

早在19xx年，就有全国人大代表联名提出议案，建议制定文化市场管理法，以解决当时存在的文化管理跟不上，文化市场立法滞后

的问题。但由于部门利益纷争、职能严重交叉等原因，文化市场管理法至今尚未形成。当前，根据以法治国的基本方略，以及大力发展文化产业的政策，要尽早推出促进文化产业大发展的法律法规，使文化产业在发展在法律层面得到保障。例如，可以出台“文化产业促进法”、“文化产业振兴与发展条例”等法律法规。

4、文化产业发展中存在着体制不顺、管理不畅等问题。 只有从提高党的执政能力建设的高度，从建设服务型政府的行政管理体制改革思路，以及党的十七大提出的大部制改革尝试中，考虑组建统筹管理我国文化产业振兴和发展的机构或部门，集中力量推进我国文化产业的发展和壮大。既可以考虑对现行的宣传文化部门设置体制进行有利于文化产业发展的组合，形成一个整合现有各部门与文化产业发展有关的职权和职能的新的综合性部门。如果组成一个负责文化产业大发展大繁荣的“大部”的条件还有待进一步成熟的话，现阶段至少也要建立一种部际联系会议制度，尽可能有效整合各部门的力量来推动文化产业的发展。

四．发展文化产业的对策建议

1、推进文化体制改革。

逐步建立和完善文化产业发展的体制机制，营造有利于出精品、出人才、出效益的良好环境。继续对国有文化企业进行公司制或股

份制改造，完善法人治理结构，建立现代企业制度，重塑市场主体，完善市场体系。

2、培育龙头企业。

形成具有规模优势的文化企业和产业布局，重点抓好文化企业上市工作。整合影视、出版、报刊、歌舞等同质优势资源，扶持文化企业进行跨地区、跨行业、跨所有制兼并重组，推动文化企业上市。

3、打造品牌。

引领文化产业发展，提升文化竞争力，把现象、亮点制成品牌、形成产业。对地区特色资源进行有效整合和深度加工。

4、培育新兴文化业态。

一是重视并扶持动漫、游戏产业发展；二是支持依法设立网络广播、网络电视，发展手机电视、移动电视、IP电视等广播电视新媒体产业；三是推动数字出版产业发展；四是加快特效电影发展；五是发展电子商务，完善网络服务功能；六是发展以研发设计、建筑设计、工艺美术设计、咨询策划、时尚设计为重点的文化设计产业，逐步形成具有吉林特色的文化创意产业。

5、建设文化产业集聚区。

建设文化产业比较聚集的地区，如少数民族聚集的地区，文化氛围较丰富的地区，有特色的文化或者建筑的地区，形成文化产业聚集区。

6、完善文化产业投融资服务体系。

一是构建完善的政策服务平台；二是组建文化产业投资公司，对重点文化产业项目建设和大型骨干文化企业上市等给予重点支持；三是建立文化产业投融资体制机制；四是加强对文化企业的贷款支持，对符合信贷条件的文化企业给予利率优惠，并积极拓展适合文化产业发展特点的贷款融资方式和相关保险服务。

五．总结

中国的经济总量和综合国力明显提升，这也为中国文化的全球性辐射奠定了物质基础。我们应加快文化创新和文化产业化的步伐，把中国建设成多姿多彩的文化出口大国，形势紧迫,不进则退。只有不断深化文化体制改革,形成具有核心竞争力的民族文化,才能增强我国文化产业的整体实力,推动中华文化走向世界,在激烈的国际竞争中立于不败之地。发展文化产业,这是中国经济社会发展的必然选择。我国文化产业近几年的业绩，已在国内外树立了良好的形象，北京申办20xx年奥运会的成功，更使国内外投资者普遍看好北京文化产业。我们需要抓住文化产业发展的机遇，创造出独具特色的东方文化产品，

与时俱进的促进我国文化产业的发展。

第二篇：文化产业1

JCultEcon(2014)38:71–84

DOI10.1007/s10824-012-9188-0

ORIGINALARTICLE

Howdoyourrivals’releasingdatesaffectyourbox

of?ce?

FernandaGutierrez-Navratil?

VictorFernandez-Blanco?LuisOrea

JuanPrieto-Rodriguez?

Received:19January2012/Accepted:3October2012/Publishedonline:20November2012?SpringerScience+BusinessMediaNewYork2012

AbstractInthispaper,westudytowhatextentamovie’sbox-of?cereceiptsareaffectedbythetemporaldistributionofrival?lms.Weproposeareduced-formempiricalmodeltomeasureandtestcompetitioneffectsamong?lmsreleasedclosetoeachotherinastandardregressionframework.Suchananalysisisappealingintermsofitspolicyimplicationandmayprovideguidancetodistributorstodecideontheirreleasingdatesoftheir?rms.Weestimatethismodelusinginformationonthe?lmsreleasedin?vecountries:theUSA,theUnitedKingdom,Germany,FranceandSpain.Thegeographicaldimensionofourdatasetpermitsustocontrolforunobservedheterogeneityamong?lmsreleasedusingpaneldatatechniques,whichallowsustoevaluatetheindividualandspeci?ceffectsofeach?lm.Thuswedealwithoneofthemostrelevantfeaturesofthemoviemarket,namelythepresenceofhighlydifferentiatedproducts.

KeywordsTemporalcompetition?Movieexhibition?Filmindustry?Paneldata?Unobservedheterogeneity?Differentiatedproducts

JELClassi?cationZ10

F.Gutierrez-Navratil(&)?V.Fernandez-Blanco?L.Orea?J.Prieto-Rodriguez

DepartamentodeEconomia,FacultaddeEconomiayEmpresa,UniversityofOviedo,

AvenidadelCristos/n,33006Oviedo,Spain

e-mail:luifergu2002@hotmail.com;gutierrezluisa@uniovi.es

V.Fernandez-Blanco

e-mail:vfernan@uniovi.es

L.Orea

e-mail:lorea@uniovi.es

J.Prieto-Rodriguez

e-mail:juanprieto@uniovi.es

123

72JCultEcon(2014)38:71–841Introduction

Themotionpictureindustryhasreceivedanincreasingattentionfromacademicsinrecentyearsduetoseveralfeatureswhichmakeitaparticularlyinterestingtopicofstudy.First,themotionpictureindustrycanbede?nedasadifferentiatedproductoligopoly.Allmoviesaredifferentbyde?nition,andthedistributionmarketisalsocontrolledbyasmallgroupofcompanies:themajor?lmstudios.Second,duetothisheterogeneityandthefactthatproductsinthisindustryareforimmediateconsumption,thelifecycleformoviesdifferssigni?cantlyfromthatoftypicalconsumerproducts,beingcharacterizedbyabox-of?cepeakinthe?rstweekofreleasethatisfollowedbyanexponentialdecaypatternovertimeasnew?lmsentertheexhibitionmarketandthevalueofthe?lmdeclines.Ingeneral,between60and70%of?lmrevenuesattheboxof?cearecollectedinthe?rst3weeks,sothatmotionpicturesexhibitionisashortlife-cycleindustry.ThesecharacteristicsareobservedbothintheUnitedStatesandtheEuropeanmarket.1Moreover,Kanzler(2011)pointedoutthatthereisatrendtowardshorterrunsandincreasedconcentrationofboxof?cearoundopeningweekend,particularlyforwidereleases.Finally,cinemademandischaracterizedby?uctuationsalongtheyear,withhighpeaksincertainweeks.2Inthiscontext,whendistributorsbelievethata?lmmaybeablockbustertheymaytrytoreleaseitinhigh-demandperiodsastheyknowthat1month’srevenueduringtheseperiodscouldproducebox-of?cesalesequivalenttoseveralmonthsinlow-demandphases.3However,whenstudiostrytomaximizebox-of?cerevenuestheyfaceaninterestingstrategicproblem,namelythetrade-offbetweentryingtocaptureasmuchoftherevenuesaspossibleduringthepeakdemandperiodsandavoidinghead-to-headcompetitionforthesameaudience.Thishead-to-headcompetitioniscriticalinthe?rstfewweeksofrunninga?lmasthewholecommerciallifeofa?lmdependsontheperformanceduringthe?rstweeks.4Moreover,thishead-to-headcompetitioncannotbealleviatedbyreducingpricesasticketpricesareverysimilarwithineachlocalmarket(seeOrbachandEinav2007;ChisholmandNorman2012),andtherearenopricedifferencesamong?lmsexhibitedinacinematheateratagivenmoment.5

Inlightoftheaforementionedfeatures,thereleasedateinthemotionpictureindustryisanessentialvariableindistributors’strategies.Thecommercialsuccessofa?lmdependscruciallyonthereleasedate,partlybecauseamovie’sopeningweekendisusuallythemostpro?tablebutalsobecausemoviesthatarereleased1

2EvidencecanbefoundinKriderandWeinberg(1998),Krideretal.(2005),andKanzler(2011).Einav(2007)foundthattheunderlyingseasonalityofbox-of?cerevenuesisampli?edbytheseasonalityinthenumberandqualityofavailablemovies.However,astheunderlyingdemandaccountsforabouttwo-thirdsoftheobservedseasonalvariationintotalsales,wewilltrytocontrolforthisunderlyingseasonalityinourempiricalapplication.

SeeMoulandShugan(2005),Vogel(2007)andRadasandShugan(1998).3

4Forinstance,agoodperformanceduringthe?rstweekendmightcreateapositiveword-of-moutheffect,capturingtheattentionofthepublic,mediaandexhibitors(DeVany2004).

Wecanobservesomepricediscrimination,butthisisintermsofdays,seasonsorconsumergroups,notamongmovies.Theexceptionisthecaseof3-Dmovies,butthesecanbeconsideredasadifferentcommodity.5

123

JCultEcon(2014)38:71–8473closetogetherarelikelytogeneratenegativeeffectsoneachother’srevenues(Corts2001).

Inthisframework,thispaperanalysestheroleoftemporalcompetitioninsidethemoviedistributionmarketanditseffectonbox-of?cereceipts.Theeconomicperformanceofmoviesisanexpandingresearch?eldinEconomics.Thereisanextensiveliteraturethatanalyseshowbox-of?cerevenuesaredeterminedbyaseriesofexplanatoryvariablesrelatedtothemovie’sclassi?cation(sequel,rating,andgenre),productionfeatures(budget,staranddirectorpower),quality(critics’reviews,awardnominationsandacademyawards),anddistributioncharacteristics(advertising,marketingexpenditureandopeningscreens).6Previouspapershavegenerallyusedmultivariatelinearregressiontopredict?lmperformancemeasuredbytwodifferentdependentvariables:cumulativedomesticrentalsandthetotallengthofrunofeach?lm.Thus,Sochay(1994)foundthatcompetition,measuredbyaconcentrationratioforaspeci?c?lm,hasasigni?canteffectonbox-of?ceperformance.Jedidietal.(1998)analyzedcompetitiveintensityduringamovie’sopeningweekendandoveritsrunandidenti?edfourdifferenttypesofmovies.ElberseandEliashberg(2003)analyzedcompetitionforscreens(i.e.,screensallocatedbyexhibitors)andforrevenues(i.e.,attentionfromaudiences)usingweeklydatafromtheUnitedStates,France,Germany,SpainandtheUK.Theydistinguishedbetweennewreleasesandongoingmoviesandfoundthatcompetitionisstrongeramongmovieswithsimilarcharacteristicsandthatthelongerthemoviesareonrelease,thelesserthecompetitioneffect.Moreover,theyobservethatcompetitionforrevenueisastrongpredictorofrevenuesthroughoutamovie’srun.Movieperformanceisalsohurtbysimultaneousreleasesofthesamegenreandrating(Ainslieetal.2005)andbythereleaseofother?lmswithasimilartargetaudience(Basuroyetal.2006).7Morerecently,Calantoneetal.(2010)haveestimatedamodelusingweeklydatathataccountsdirectlyandindirectly(throughthenumbersoftheaters)forcompetitiveeffectsontheperformanceofamovie.Theyconcludethatongoing(i.e.,incumbent)movieshaveahighernegativeeffectthannewreleases.Thisliteraturecon?rmstheimportanceofconsideringcompetitionasakeyfactorindeterminingbox-of?cerevenues.Ourpaper,ontheotherhand,dealsexplicitlywiththeeffectsofcompeting?lmsreleasedinthepast,presentandfutureweeksonthetotalbox-of?cerevenuesofaparticular?lm.Weusepaneldatatechniquesandthuswecontrolforthespeci?ceffectsofeach?lm.Indoingso,wewilltakeintoaccountthat‘‘LittleMissSunshine’’isnot‘‘Avatar’’.

Althoughothersourcesofrevenuesmightbemoreimportant,8theatersarearelevantmarketnotonlyintermsofrevenue?guresbutalsoasapredictoroftherevenuesofa?lminothermarkets.Forinstance,McKenzie(2010)concludedfortheAustralianmarketthatsuccessattheboxof?ceisanextremelyimportantdeterminantofthedemandforDVDs.Additionally,Langetal.(2011)pointedoutthatpreviousbox-of?cesuccesshasstrongpositiveeffectsonDVDsales.Thepast6

7SeeHadida(2009)andMcKenzie(2012)forasummaryofthisliterature.However,Basuroyetal.(2006)havefoundthatongoing?lms’competitionforscreenshasapositiveeffectonweeklybox-of?cereceiptsandscreencoverage.

SeeGinsburghandWeyers(1999),Ravid(1999),Hennig-Thurauetal.(2006).8

123

74JCultEcon(2014)38:71–84box-of?ceperformanceofamovieappearstobethesinglemostimportantdeterminantofDVDsales.

Consequently,ourempiricalmodelusesaggregatemovietheaterrevenues(foreach?lmineachcountry),andhence,itcanbeviewedasa‘‘reducedform’’ofmorecomplexmodelsbasedonweeklydata.Ourempiricalstrategyallowsustomeasuretheeffectsofcompetitionontotalbox-of?cerevenueinadirectmanner,withouttheneedtodealwitheconometricissuesthatappearwithdynamicmodels.Forinstance,theword-of-moutheffectnotonlyrequirestheuseofweeklydatabutalsotheestimationofdynamicmodelswithdifferentsourcesofendogeneityandautore-gressiveerrors.Thisdynamicframeworkmakesitdif?culttocontrolforunobservable?lmeffectsthatarelikelycorrelatedwithsomeofourexplanatoryvariables(seeVerbeek2008).However,ifunobservedheterogeneityisnotaddressed,wewillhavebiasedestimators.Sinceunobservedheterogeneityisexpectedtobeimportantinourapplication,adynamicmodelmaythereforenotbethemostappropriate.Furthermore,althoughtheweekly-basedmodelscanbeusedtomeasurethecompetitioneffectontotalrevenues,thiswouldrequirecomputingasetofderivativesandobtainingcon?denceintervalsfortheestimatedeffectsisnotstraightforwardinadynamicsetting.Byestimatingareduced-formmodel,however,wecanmeasuretheeffectsofcompetitionontotalbox-of?cerevenueinadirectway,givingpractitionersasimplemanagementtool.Moreover,areduced-formmodelallowsustocontrolfortheindividualeffectsspeci?ctoeach?lmandtheendogeneityproblems.

Ourreduced-formspeci?cationprovidesasimpleeconometricspeci?cationdesignedspeci?callytomeasureandtestcompetitioneffectsontotalbox-of?cerevenuesof?lmsreleasedcloseeachother.Correctinferenceabouttheseeffectscanbecarriedoutinastandardregressionframeworkusingthewell-known?xedeffectsestimator.Asthecoef?cientsofourmodelhavesimpleinterpretations,themodelisausefulpolicytoolwhichcanbeusedtoprovideguidancetodistributorstodecidethereleaseoftheir?lms.

Moreover,ourempiricalmodelallowsustobothmeasureandtesttowhatextentthepast,presentandfuturerivalreleaseshaveasymmetriceffectsonthetotalbox-of?cerevenuesofaparticular?lm.Thisinformationisofcrucialrelevanceto?lmstudiosanddistributorsbecausetheyoftencarryoutintensemarketresearchbeforereleasingtheirmoviesinordertodiscoveraudience’spreferencesandanticipatemarketresponses.Ourempiricalresultscouldhelpthemtousetheirreleasedatesasastrategicvariabletokeeporcapturemarketsharefromtheirrivals.Withthisinformation,thedistributorscouldimprovereleasetimingdecisionsthroughknowledgeofthepayoffmatrixofthereleasegame,comparingthegainsofavoidingcompetitionbyopeningbeforearivalorbydelayingthereleaseandopeninglater.

Thepaperisorganizedasfollows.Inthenextsection,wedescribethedatabaseused,andinSect.3,theempiricalmodeltobeestimatedisoutlined.InSect.4,theprincipalresultsarepresented.The?nalsectionconcludes.

123

JCultEcon(2014)38:71–84752Sampleanddatabase

Thissectionsummarizesthedataweusedtoperformourempiricalanalysis.Thesampleconsistsofbox-of?cerevenuesformoviesreleasedin?vecountriesfromJanuary1st,2000toDecember31st,2009.WehavecollectedtheinformationprovidedbyA.C.NielsenEDIonmoviesreleasedintheUnitedStatesandthefourlargestEuropeanmotionpicturesmarkets(UnitedKingdom,France,GermanyandSpain).Fromthisinformation,wehaveselectedthosemoviesreleasedinatleastthreeofthesecountries.Our?naldatabasecomprises2,811moviesand11,908validobservations.Thisdatabasehasapaneldatastructurewherethetimedimensionhasbeensubstitutedbyaspatial(country)dimension.Inordertoarriveatthisstructure,we?rsthadtomatchthemoviesacrosscountriesbecausemanymovieswerereleasedwithdifferenttitlesineachcountry.ThiswasdonebyusingtheinformationprovidedbytheInternetMovieDatabasewebsite().

Foreachmovieandcountry,ourdatasetincludesthefollowinginformation:thecorrespondingtitles,theof?cialreleasedates,totalbox-of?cerevenues,numberoftheatersonthereleasedate,maximumnumberoftheaterscountedfora?lmoverthecourseofitsrun,thedistributors,andtheMPAArating.InthecaseofFrance,wehavetotalattendanceinsteadoftotalbox-of?cerevenues.

3Theempiricalmodel

Accordingtothepreviousliterature,thebox-of?cerevenuesofa?lmdependonthecharacteristicsthatdeterminethequalityofthe?lm,thecompetitiveenvironmentandtheseasonalityinunderlyingdemand.9Consideringallthesedeterminants,weproposeestimatingthefollowingempiricalmodel:

MarketShareic?ai?cfMaxThtsic?cffMaxThts2ic

?3X

r??352Xw?2?krRivThts?r?ic?krrRivThts?r?2ic??hwD?w?ic?eic?1?

wheresubscriptcstandsforcountry,subscriptwidenti?es?lmi’sopeningweek,andeicistheerrorterm.The?lm-speci?cconstanttermaicapturesobservedandunob-served?lmcharacteristics.Somecharacteristicssuchasgenre,awards,stars,dis-tributor,etc.mightbeavailablebutcannotbeincludedina?xedeffectspaneldatamodel.Therearemanycharacteristicsthatareunobservable,andthesemaybeimportantbecausethemotionpictureindustryisoneofthemosthighlyproduct-differentiatedmarkets,aseachmovieisuniquebynature.Thisiscon?rmedbyEinav(2007),whofoundthat‘‘observablevariables(…)explainasmallfractionofthevariationinquality’’.Assomeoftheseunobservedcharacteristicsarelikelytobecorrelatedwithourexplanatoryvariables(e.g.,theatersnumbers),ourparameterestimateswillbebiased.ThisproblemappearsifweestimatethemodelbyOLS9?rrez-Navratiletal.(2011).Formoredetails,seethetheoreticalmodeldevelopedinGutie

123

76JCultEcon(2014)38:71–84whetherwehavecrosssectionalorpaneldata.However,ifwehaveapaneldataset(asisthecaseinourapplicationwherethetimedimensionhasbeensubstitutedbyacountrydimension),wecandealwiththisproblemusinga?xedeffectsestimatorthatcontrolsfor?lmcharacteristicsthatcanbeconsideredinvariantacrosscountries.ThedependentvariableMarketShareicisthelogof?lmi’stotalbox-of?cerevenuesineachcountry,normalizedbythecountry’sannualbox-of?cerevenues.Thisnormali-zationallowsustocontrolforthecountry-sizeeffects,changesincinemarevenuesovertime,in?ationandchangesintherelativepricesofmovieticketsduringtheperiod.10Followingthepreviousliterature,asexplanatoryvariables,weincludetheopening-weektheatersofthemoviei,OwnThtsic,toproxythepotentialattractivenessofthe?lm,thatis,itsex-antecompetitiveintensityorpower(seeHadida2009).Sincetherelevantgeographicalmarketinthemovietheaterindustryislocalinnature(Davis2006;Sunada2012),thevariableOwnThtsicmayalsobeviewedasan‘‘inverse’’measureofcustomertransactioncosts.Forbothreasons,weexpectapositivecoef?cientforthisvariable.

NeelameghamandChintagunta(1999)pointedoutthatthenumberofscreensmightbeconsideredendogenousbecausetheevolutionofthenumberoftheatersovertimewillpartiallydependontheperformanceofthemovie.Hence,asourdependentvariableaggregatesallweeklyrevenues,anymeasureofthenumberofscreensallocatedtoamoviebeyondtheopeningweekislikelyanendogenousvariable(seeElberseandEliashberg(2003)formoredetailsaboutthisinterdepen-denceorsimultaneity).However,exhibitorsallocatescreenstoamovieinitsopeningweekbasedontheirexpectationsregardingpotentialdemand,whichdoesnotdepend,byde?nition,onpreviousrevenues.Instead,theexpectedopening-weekrevenuesdependonthecharacteristicsofthe?lms.Asthefullsetofcharacteristicsof?lmsisbeingcontrolledbyourmovie-speci?ceffects,ourestimatorsolvesthissourceofendogeneityregardingopening-weektheaters.

Toexamineindetailthecompetitioneffectof(past,presentandfuture)rival?lmsontotalbox-of?cerevenues,weincludeasetofvariables,RivThts(r)ic,thatmeasurestheopening-weektheatersofalltherivalsofthemovieireleasedthesameweek(inthiscase,r=0),upto3weeksbefore(-3Br\0)and3weeksafter11(0\rB3)thereleaseofmoviei.Thisvariabletakesintoaccountboththenumberofrivalsandtheirabilitytocaptureattendance.Alltheseindependentvariableswerenormalizedbysubtractingtheirmeans,sothe?rst-ordercoef?cientscanbeinterpretedasderivativesevaluatedatthesamplearithmeticmean.Sincepreviousliteraturehasfoundthatcompetitioneffectsmaydependon?lmi’scharacteristicswesplitthesampleusingtheMPAAratingandestimateEq.(1)forthethreedifferentgroupsof?lms:general,teenagerandrestrictedaudiences.12

Duetodatalimitations,themarketshareinFrancehasbeencomputedusingtotalattendanceinsteadoftotalbox-of?cerevenues.

Wefocusouranalysisonthemostpotentiallyharmfulrival?lms,thosereleasedduringtheperiodfrom3weeksbeforeto3weeksaftermoviei’srelease.

Wejustusedthreecategoriessincewehadtoharmonizeratingscalesthatdifferacrosscountries.Thegeneralaudiencegroupincludes?lmssuitableforallagegroupsandforchildrenover6years(GandPGrating).Therestrictedaudiencegroupincludes?lmsmoresuitableforagesover17years(Rrating)andtheteenageraudiencesgroupincludes?lmssuitableforteenagers(PG-13).121110

123

JCultEcon(2014)38:71–84Table1SummarystatisticsofdataDataMarketshareOwntheatersRivals’theaters0Rivals’theaters1Rivals’theaters2Rivals’theaters3Rivals’theaters-1Rivals’theaters-2Rivals’theaters-3

UnitsPercentageThousandsThousandsThousandsThousandsThousandsThousandsThousandsThousands

Mean0.0036730.4221.8162.0342.0452.0322.0612.0362.047

Min1.75E-080.0010.0000.0000.0000.0000.0000.0000.000

Max0.1133054.36615.38715.38815.38815.38815.38815.38815.388

SD

77

0.0072050.8032.4262.5992.6002.5662.6112.5512.597

Note,inaddition,thatwealsoincludesquaredvaluesofbothOwnThtsicandRivThts(r)ictocapturenonlinearsizeandcompetitioneffects,respectively.Finally,followingthesamestrategyasEinav(2007)toestimateseasonalityinunderlyingdemand,weincludeasetofweeklydummyvariables(the?rstweekissetasthebase).AsummaryofthedescriptivestatisticsfortheabovevariablesisshowninTable1.

4Estimationandresults

Equation(1)canbeviewedasapaneldatamodel.Asiscustomary,itcanbeestimatedusingeithera?xedeffectsorarandomeffectsestimator.13Table2displaystheresultsofboth?xedeffectsandrandomeffectsestimations.Weprovidecluster-robuststandarderrors,whereclusteringby?lmpermitscorrelationoftheerrorswithin?lmsbutconstrainserrorstobeindependentacross?lms.OnthebasisoftherobustHausmantest,14werejectthenullhypothesisthattheindividualeffect(ai)andregressorsareuncorrelated.15Thus,the?xedeffectsmodelisourpreferredmodelasitallowsustoexplainmarketsharescontrollingforunobserveddifferencesacrossmovies.

TheresultsinTable2showthattheFEmodelexplainsnearly29%ofthemarketsharevariance.Mostoftheestimated?rst-ordercoef?cientshavetheirexpectedsignsatthesamplemean.The?rst-ordercoef?cientoftheopening-weektheaterswherethe?lmwasexhibited(OwnThtsic)ispositiveandstatisticallysigni?cant,asexpected.ThisresultisinlinewithCalantoneetal.(2010)andElberseandEliashberg(2003),thoughweprovideevidenceofadecreasingeffectofOwnThtsicastheestimatedsecond-ordercoef?cientofthisvariableisnegativeand

WeuseanFtesttotestwhetherthe?lm-speci?cconstantterms(ai)areallequal.Weobtainavalueof5.23,whichfarexceedsanycriticalvalueofanF(2810,9030).Therefore,werejectthenullhypothesisinfavoroftheindividualeffectsmodel.

141513

WefollowCameronandTrivedi(2009),whousethemethodofWooldridge(2002).

ThevaluefortheF(16,2810)statisticis115.64,aboveanycriticalvalue,sowestronglyrejectthenullhypothesisthattheindividualeffectsandexplanatoryvariablesareuncorrelated.

123

78

Table2FixedeffectsandrandomeffectsmodelVariable

ln(MarketShare)FixedeffectsmodelCoef?cient

OwnThtsRivThts0RivThts1RivThts2RivThts3RivThts-1RivThts-2RivThts-3(OwnThts)

2

JCultEcon(2014)38:71–84

Randomeffectsmodel

Robust-t40.52-4.62-3.40-3.17-1.27-3.11-2.06-0.42-29.873.881.213.120.781.802.450.18-66.18

Coef?cient3.6606***-0.2005***-0.1002***-0.0956***-0.0444*-0.0822***-0.0332-0.0066-0.0010***2.2E-05***6.0E-06**1.0E-05***3.0E-065.7E-06**4.8E-06*-4.5E-08Yes-6.9967***11,908

-66.96Robust-t60.17-8.52-4.27-4.05-1.82-3.32-1.39-0.27-38.577.532.323.961.032.081.78-0.02

2.7530***-0.1010***-0.0750***-0.0709***-0.0298-0.0734***-0.0473**-0.0099-0.0007***1.1E-05***3.0E-067.6E-06***2.2E-064.7E-06*6.3E-06**4.5E-07Yes-6.8914***11,9080.284836.470

F(16,2810)=115.64Prob[F=0.0000

(RivThts0)2(RivThts1)(RivThts2)(RivThts3)

2222

(RivThts-1)(RivThts-3)ConstantNR2

(RivThts-2)2

2

Weeklydummies

F(67,2810)Hausmantest

***Signi?cantat1%level;**signi?cantat5%level;*signi?cantat10%level

statisticallysigni?cant.Hence,increasingthenumberoftheatersisfoundtoraisethebox-of?ceperformancebutnotinde?nitely.

Generallyspeaking,theremaining?rst-ordercoef?cientsshowthatthepresenceofmore(orstronger)past,contemporaneousandfuturecompetitorshasastatisticallysigni?cantandnegativeimpactonmarketshare.Thepositivevaluesofthesecond-ordercoef?cientsindicate,however,thatthecompetitioneffectisdecreasingwiththenumberofrivals.Regardingthetemporalpatternoftheestimatedcompetitioneffects,wehavefound,asexpected,thatcontemporaneousrivalmovieshaveastrongereffectthan?lmsreleasedinotherweeks.Thiseffectincreasesasrival?lmreleasescomecloser.TheresultsinTable2showevidenceofanasymmetrybetweentheeffectofpastandfuture?lmreleases.Thetemporalpatternofcompetitioneffectswillbediscussedinmoredetaillaterinthissection.Wehaverecoveredalloftheestimatedmovie?xedeffectstochecktherobustnessofouranalysis.Weobtainedaleft-skeweddistributionwitharelatively

123

JCultEcon(2014)38:71–8479largevariance.Accordingtotheseresults,ifamovieisareal?asco,thereseemstobenolimittoitsfailure.Ontheotherhand,wehavefoundthatAmelie,AmericanBeauty,Avatar,BillyElliot,BrokebackMountain,Chicago,Chocolat,CrouchingTigerHiddenDragon,TheDarkKnight,FindingNemo,GranTorino,HarryPotterI,II,IIIandIV,IceAgeIII,TheIncredibles,LostinTranslation,MillionDollarBaby,MyBigFatGreekWedding,PiratesofCaribbeanIIandIII,Ratatouille,Shrek2,SlumdogMillionaireandToyStoryIIarethe?lmslocatedabovethe99thpercentilevalueinthe?xedeffectdistribution.

Previousliteraturesuggeststhatthecompetitioneffectstronglydependsonmovies’characteristics,includingtheirrating.As?lmswithdifferentratingsareaimedatdifferenttargetaudiences,onewouldexpectasubstantialdifferenceinthewaythetemporalcompetitionaffectstotalbox-of?cerevenuesof?lmswithdifferentratings.Toexaminethisissue,wehaveestimatedourmodelforthreegroupsofmovies:general,restrictedandteenageraudience.Foreachratingtype,weconsiderallrivalmoviesregardlessoftheirrating.TheparametersestimatedareshowninTable3.

Thethreemodelsestimatedhavesimilargoodnessof?t,withthegeneralaudiencemodelexplaining28%ofthemarketsharevariance,themodelfortherestrictedaudienceexplaining36%whilethe?gureis30%fortheteenageraudiencemodel.Aswiththeoverall-samplemodelinTable2,theHausmantestallowsustorejecttheRandomEffectsestimatorineachmodel.Again,mostoftheestimatedcoef?cientsretaintheexpectedsigns.Theopening-weektheaterswhereamoviewasexhibited(OwnThtsic)stillhaveapositiveandstatisticallysigni?canteffectonthemarketshareinallmodels.Thiseffectisagainsigni?cantlydecreasing.

Regardingthecompetitioneffect,theresultsarenotasclearasthoseshowninTable2becauseboththemagnitudeandstatisticalsigni?canceoftheestimatedparametersvaryacross?lmgroups.Ingeneral,therival?lmsthathaveasigni?cantimpactareonlythosereleasedclosetothe?lm.Therivalmoviesreleasedthesameweekhaveasigni?cant,negativeanddecreasingeffectinallcases.Inthecaseofgeneralaudience?lms,otherharmfulcompetitorsarethosethatwerereleased1weeklater.Forrestrictedaudience?lms,themostimportantrivalsarethosereleasedthepreviousweek.Inthecaseofteenageraudience?lms,rivalsreleased2weekslaterhaveasigni?canteffect.

Inordertocomparethemagnitudesofthedifferentcompetitioneffectsinaclearerway,wehavecalculatedtheelasticitiesofthemarketsharestakingintoaccountthelinearandsquaredeffects(seeTable4).Allelasticitiesareevaluatedatthemeanofeachgroupofmovies.Inthistable,weusetheelasticityofthemarketsharewithrespecttoOwnThtsicasabenchmarktodiscussthecompetitioneffects.Inallmodels,theestimatedelasticitiesrelativetotheopening-weektheatersarealwayshigherthanunity.Thisindicatesthatthereareeconomiesofscaleatthemean,thatis,youcantakeadvantageofthesizeofthereleasestosomeextent,butthenegativequadraticeffectsindicatethatitwilldecreaseasthe?lmisreleasedinmoretheaters.

Regardingthecompetitioneffects,themostharmfulrival?lmsinallmodelsarethose?lmsreleasedthesameweek.However,theelasticityofthemarketsharewith

123

80JCultEcon(2014)38:71–84

Table3Fixedeffectsmodelwithmoviesforgeneral,restrictedaudienceandteenagersVariable

ln(MarketShare)FE(generalaudience)Coef?cient

OwnThtsRivThts0RivThts1RivThts2RivThts3RivThts-1RivThts-2RivThts-3(OwnThts)

2

FE(restricted)Coef?cient2.3132***-0.1750***-0.09100.0186-0.1277**-0.1310**-0.00010.1141*-0.0007***1.6E-05***7.4E-062.6E-061.7E-05***6.4E-06-9.0E-07-1.3E-05**Yes-7.09645***2,1930.3620

F(67,1400)=5.22F(16,1400)=12.15Prob[F=0.0000

Robust-t12.88-3.54-1.420.28-1.96-2.110.001.95-10.092.780.970.372.620.96-0.14-2.12-30.95

FE(teenagers)Coef?cient2.6076***-0.0596*-0.0858**-0.0788**-0.0184-0.0368-0.0606-0.0306-0.0007***4.0E-065.9E-078.1E-06**-4.8E-07-5.2E-077.6E-06*1.6E-07Yes-6.87982***5,4120.3041

F(67,2362)=15.05F(16,2362)=47.55Prob[F=0.0000

-38.84Robust-t25.34-1.76-2.43-2.27-0.48-0.99-1.61-0.85-18.480.930.162.21-0.10-0.121.810.04

Robust-t17.70-4.24-2.61-0.32-1.22-1.23-0.45-0.68-16.183.881.360.402.480.991.221.70-29.45

2.8366***-0.1937***-0.1089***-0.0154-0.0555-0.0599-0.0204-0.0333-0.0007***2.4E-05***6.5E-062.3E-061.4E-05**5.9E-066.6E-068.5E-06*Yes-7.26028***4,3070.2759

(RivThts0)2(RivThts1)(RivThts2)(RivThts3)

2222

(RivThts-1)(RivThts-3)ConstantNR2F

(RivThts-2)2

2

Weeklydummies

F(67,2150)=7.99F(16,2150)=38.76Prob[F=0.0000

Hausmantest

***Signi?cantat1%level;**signi?cantat5%level;*signi?cantat10%level

respecttocontemporaneousrivalsisalwayslowerinabsolutetermsthanone.Thiselasticityforallmoviesindicatesthat,atthemean,a1%increaseincontempo-raneousrivaltheatersreducesbox-of?ceshareby0.18%.Thiseffectisevenmoreimportantforrestrictedaudience?lmswherethecorrespondingelasticityistwiceasthatfoundintheoverallmodel(about0.41).UsingtheelasticityofthemarketsharewithrespecttoOwnThtsicasabenchmark,weconcludethatthecontemporaneouscompetitioneffectrepresents,inabsoluteterms,lessthana?fthoftheown-effectoftheatersfortheoverallmodel.Thiseffectgoesuptoone-thirdforrestrictedaudience?lmsanddecreasesmarkedlyfor?lmsnotrestrictedtoteenagers.

Inthecaseofthemodelthatconsidersallmovies,wheretheeffectsofbothpastandfuturerivalsaresigni?cantandnegative,wegetadecreasingimpactaswemoveawayovertime,thatis,theclosertherelease,thehigherthenegativeimpact.ItisalsoworthmentioningthattheestimatedelasticitiesinTable4showtheexistenceofasymmetriceffectsofpastandfuture?lmreleases.Inthismodel,we

123

38:71–84

43280977344098o9643705i0110010t.......a0000000R-------)s*re***g****a73340491n16878694e90753716e01110010t........(10000000g-------

54691674919883o6082403i3202302t.......a0000000R-----)d**e****tc*****i43029061rt77416991s10395806e14202302r........(10000000g-----

03790114428454o4428835i2100000t.......a0000000R-------

lev

)elecn%ei0d1uta***al***ta***nr33314335ean2934497765529835ce?g02100000i

........n(10000000gg-------i

s*;lev

1919025e7131017lo5325381i1110100%t.......)a00000005XR-------t

a

ntl(nqan/c

)a?evime*****nR******g******iteara04435221s

80119930*

dh)eSl48545492*

l111101tana00........;

l(10000000lu(lqg--a-----evevle%se1iti0123123tc---aissssssstsrrrrrrrtaseeeeeeenlrtttttttaEeaaaaaaac

teeeeeeeahhhhhhh?

i

4eetttttttn

lh’’’’’’’gebtsssssssilabinlllllll

raaaaaaaSaawvvvvvvv*TVOiiiiiiiRRRRRRR*

*81

123JCultEcon(2014)

82JCultEcon(2014)38:71–84canseethatfuturerivalsalwayshaveahighereffectthanpastrivals.Indeed,theeffectofpastrivalsdecreasesmorerapidlyasthereleaseisfartheraway.

Comparingtheestimationsofthedifferentratingmoviegroups,differentpatternswithrespecttotheimpactofpreviousorsubsequentreleasesaredetected.Marketshareismainlyaffectedbycompetitorslaunchedbeforetheownreleaseforrestricted?lmsandthoselaunchedafterfortheothertwogroupsofmovies.Furthermore,theimpactscausedbynon-contemporaryrivalsareonlycomparabletocontemporaryrivalsinthecaseofrestricted?lms.

Summarizingallthese?ndings,itwouldappearthattheaudienceofnon-restricted?lmsismoreinterestedinnoveltysincetheyaremoreaffectedbythose?lmsthatwillbereleasedinthefollowingweeks.However,therestrictedmovies’audienceappearstobemoresensitivetoaword-of-moutheffectrelatedtopreviouslyreleasedpictures.Consequently,itseemsthatthereleasingpolicyofrestrictedaudience?lmsshouldpaymoreattentiontorival?lmsalreadyonexhibition,whereasthedistributionofothermoviesshouldpaymoreattentiontofuturerivals.

5Conclusionandmanagerialimplications

Inthispaper,weshowtheroleoftemporalcompetitioninthemovieexhibitionmarketanditseffectonmarketshare.Wetakeintoaccountthepossibledifferencesbetweenpresent,pastandfuture?lmreleasestotesttheextenttowhichtemporalcompetitionhasasymmetriceffects.Informationabouttheexistenceandthedetailsoftheseasymmetriesareofrelevanceformanagerialdecisionmaking.Theycanusethisinformationtomanagetheirreleasedatestodefendorcapturemarketsharefromtheirrivals.

Weproposeanempiricalapplicationtoexamineourmainresearchquestion.Weusetheinformationonthe?lmsreleasedin?vecountries,wherethegeographicaldimensionallowsustousepaneldatatechniquestocontrolfortheunobservedheterogeneityofthereleased?lmsandhencecaptureoneofthemostrelevantfeaturesofmoviemarket:thepresenceofhighlydifferentiatedproducts.Wespeci?callyconsiderthecompetitorslaunchedinthe3-weekintervalaroundthereleasedateofaparticularmovie.Wealsoestimatethreeadditionalmodelsbysplittingthesampleof?lmsaccordingtotheirratingstocheckfordifferentialeffectsbytypeof?lm.

Inthemainmodel,whichconsidersallmovies,wefoundthattheeffectofcontemporaryrivalsisalwayslargerthanthatofprevious-releasedorfuturerivals.Ingeneral,we?ndadecreasingimpactofother?lmreleasesastheirlaunchdatesmoveawayintimefromthereleaseweekofthereference?lm.Regardingthetemporalpatternofcompetitioneffects,weobservethatfuturerivalsalwayshaveahighereffectthanpastrivals.Infact,theeffectofpastrivalsdecreasesmorerapidlyasthedistancebetweenreleasesgrows.Thisresultshowsevidenceofaclearasymmetrybetweentheeffectofpastandfuture?lmreleases.Suchasymmetryshouldbeconsideredbymanagerswhenmakingdecisionsaboutthechoiceofreleasedates.Ingeneral,itseemsmoreimportanttoavoidcompetitionfromfuture123

JCultEcon(2014)38:71–8483releasesthanfrommoviesalreadyonscreen.Thus,ifthereisapotentialblockbustertobereleased,itwouldbepreferabletoreleaseour?lmlaterratherthansooner.Inanycase,itisnecessarytoconsiderthetypeofaudienceaimedatwhentakingthesedecisions.Whenwecomparethethreeadditionalmodelsaccordingtomoviesratings,wefoundsubstantialdifferencesamongthemanddifferentpatternswithrespecttotheimpactofpreviousorsubsequentreleases.Accordingtotheabove,thereleasingpolicyshouldbedifferentforthedifferenttypesof?lms.

Regardingthegeneralaudiencemovies,thesearemoreaffectedbyrivalmoviesreleasedthesameweek,andotherimportantrivalsarethosereleasedthefollowingweek.Theaudienceofmoviesforthegeneralpublicismoreinterestedinnovelty.Anexpectedblockbusterwillleadgeneralaudience?lmstobepostponed.However,therestrictedmoviesaremoreaffectedbyrivalsthatwerereleasedthesameweekortheweekbefore.Consequently,therestrictedmoviesaudiencemaybemoresensitivetoaword-of-moutheffectrelatedtopreviouslyreleasedpictures.Ifthereisanimportantrival,restrictedmoviesshouldthereforeanticipaterivals‘releaseandbelaunchedbeforethem.

Theteenageraudiencemoviesaremoreaffectedbyrivalsreleasedinthefollowing2weeksthanbythoserivalsreleasedthesameweek.Thisaudiencealsoseemstobemoreinterestedinnovelty,aswithgeneralpublic?lms.Itisthereforebettertoavoidcompetitionbydelayingthereleaseandopeninglaterthanotherblockbusters.

Theseresultsshouldprovidesomeguidancetodistributorstoimprovetheirreleasetimingdecisions,usingthereleasedateasastrategicvariabletomaximizetotalbox-of?cerevenues.Sincethechoiceofreleasedateissensitivetothekindofproduct,itisrecommendedthatdistributorsdiversifytheirportfoliosofmovies.AcknowledgmentsThisresearchhasbeenfundedwithsupportfrom:theDepartmentofEducation,UniversitiesandResearchoftheBasqueGovernment;theEuropeanCommission(EUCultureProgrammeproject#2012-0298/001);theGovernmentofSpain(projects#ECO2011-27896,#ECO2010-17590and#ECO2010-17240).Itre?ectstheviewsonlyoftheauthors,andtheCommissioncannotbeheldresponsibleforanyusewhichmaybemadeoftheinformationcontainedtherein.References

Ainslie,A.,Dreze,X.,&Zufryden,F.(2005).Modelingmovielifecyclesandmarketshare.Marketing

Science,24(3),508–517.doi:10.1287/mksc.1040.0106.

Basuroy,S.,Desai,K.K.,&Talukdar,D.(2006).Anempiricalinvestigationofsignalinginthemotion

pictureindustry.JournalofMarketingResearch,43(2),287–295.doi:10.1509/jmkr.43.2.287.

Calantone,R.J.,Yeniyurt,S.,Townsend,J.D.,&Schmidt,J.B.(2010).Theeffectsofcompetitionin

shortproductlife-cyclemarkets:Thecaseofmotionpictures.JournalofProductInnovationManagement,27(3),349–361.doi:10.1111/j.1540-5885.2010.00721.x.

Cameron,A.C.,&Trivedi,P.K.(2009).MicroeconometricsusingStata.Texas:StataPress.

Chisholm,D.,&Norman,G.(2012).Spatialcompetitionandmarketshare:Anapplicationtomotion

pictures.JournalofCulturalEconomics,36(3),207–225.

Corts,K.S.(2001).Thestrategiceffectsofverticalmarketstructure:Commonagencyand

divisionalizationintheUSmotionpictureindustry.JournalofEconomics&ManagementStrategy,10(4),509–528.doi:10.1162/105864001753356088.

Davis,P.(2006).Spatialcompetitioninretailmarkets:Movietheaters.TheRandJournalofEconomics,

37(4),964–982.

123

84JCultEcon(2014)38:71–84DeVany,A.S.(2004).Hollywoodeconomics:Howextremeuncertaintyshapesthe?lmindustry

contemporarypoliticaleconomyseries.Routledge:Taylor&Francis.

Einav,L.(2007).SeasonalityintheU.S.motionpictureindustry.TheRandJournalofEconomics,38(1),

127–145.

Elberse,A.,&Eliashberg,J.(2003).Demandandsupplydynamicsforsequentiallyreleasedproductsin

internationalmarkets:Thecaseofmotionpictures.MarketingScience,22(3),329–354.doi:10.1287/mksc.22.3.329.17740.

Ginsburgh,V.,&Weyers,S.(1999).Ontheperceivedqualityofmovies.JournalofCulturalEconomics,

23(4),269–283.doi:10.1023/a:1007596132711.

?rrez-Navratil,F.,Ferna?ndez-Blanco,V.,Orea,L.,&Prieto-Rodriguez,J.(2011).HowdoyourGutie

rivals’releasingdatesaffectyourboxof?ce?ACEIWorkingPaperSeries.AWP-4-2011,theAssociationforCulturalEconomicsInternational.

Hadida,A.L.(2009).Motionpictureperformance:Areviewandresearchagenda.InternationalJournal

ofManagementReviews,11(3),297–335.doi:10.1111/j.1468-2370.2008.00240.x.

Hennig-Thurau,T.,Houston,M.B.,&Walsh,G.(2006).Thedifferingrolesofsuccessdriversacross

sequentialchannels:Anapplicationtothemotionpictureindustry.JournaloftheAcademyofMarketingScience,34(4),559–575.doi:10.1177/0092070306286935.

Jedidi,K.,Krider,R.,&Weinberg,C.(1998).Clusteringatthemovies.MarketingLetters,9(4),393–405.

doi:10.1023/a:1008097702571.

Kanzler,M.(2011).TheEuropeanCinemaMarkets2010inFigures.EuropeanAudiovisualObservatory.

Presentationintheframeworkofthe17theditionoftheSarajevoFilmFestival.http://www.obs.coe.int/online_publication/expert/europeancinemamarkets2010.pdf.en.

Krider,R.E.,Li,T.,Liu,Y.,&Weinberg,C.B.(2005).Thelead-lagpuzzleofdemandanddistribution:

Agraphicalmethodappliedtomovies.MarketingScience,24(4),635–645.doi:10.1287/mksc.1050.0149.

Krider,R.E.,&Weinberg,C.B.(1998).Competitivedynamicsandtheintroductionofnewproducts:

Themotionpicturetiminggame.JournalofMarketingResearch,35(1),1–15.

Lang,D.M.,Switzer,D.M.,&Swartz,B.J.(2011).DVDsalesandtheR-ratingpuzzle.Journalof

CulturalEconomics,35(4),267–286.

McKenzie,J.(2010).Howdotheatricalboxof?cerevenuesaffectDVDretailsales?Australianempirical

evidence.JournalofCulturalEconomics,34(3),159–179.

McKenzie,J.(2012).Theeconomicsofmovies:Aliteraturesurvey.JournalofEconomicSurveys,26(1),

42–70.doi:10.1111/j.1467-6419.2010.00626.x.

Moul,C.C.,&Shugan,S.M.(2005).Theatricalreleaseandthelaunchingofmotionpictures.InC.

C.Moul(Ed.),Aconcisehandbookofmovieindustryeconomics(pp.80–137).Cambridge:CambridgeUniversityPress.

Neelamegham,R.&Chintagunta,P.(1999).ABayesianmodeltoforecastnewproductperformancein

domesticandinternationalmarkets.MarketingScience,18(2),115–136.

Orbach,B.Y.,&Einav,L.(2007).Uniformpricesfordifferentiatedgoods:Thecaseofthemovie-theater

industry.InternationalReviewofLawandEconomics,27(2),129–153.doi:10.1016/j.irle.2007.06.002.

Radas,S.,&Shugan,S.M.(1998).Seasonalmarketingandtimingnewproductintroductions.Journalof

MarketingResearch,35(3),296–315.

Ravid,S.A.(1999).Information,blockbusters,andstars:Astudyofthe?lmindustry.Journalof

Business,72(4),463–492.

Sochay,S.(1994).Predictingtheperformanceofmotionpictures.JournalofMediaEconomics,7(4),

1–20.

Sunada,M.(2012).Competitionamongmovietheaters:AnempiricalinvestigationoftheToho–Subaru

antitrustcase.JournalofCulturalEconomics,36(3),179–206.

Verbeek,M.(2008).Aguidetomoderneconometrics(3rded.).NewYork:Wiley.

Vogel,H.L.(2007).Entertainmentindustryeconomics.Aguidefor?nancialanalysis(7thed.).

Cambridge:CambridgeUniversityPress.

Wooldridge,J.M.(2002).Econometricanalysisofcrosssectionandpaneldata.Cambridge,MA:MITPress.

123