Handbook of Industrial Organization, Volume 5 Edited by Kate Ho Department of Economics Princeton University Princeton, NJ, United States Ali Hortaçsu The Kenneth C. Griffin Department of Economics University of Chicago Chicago, IL, United States Alessandro Lizzeri Department of Economics Princeton University Princeton, NJ, United States North-HollandisanimprintofElsevier Radarweg29,POBox211,1000AEAmsterdam,Netherlands TheBoulevard,LangfordLane,Kidlington,OxfordOX51GB,UnitedKingdom Copyright©2021ElsevierB.V.Allrightsreserved. Nopartofthispublicationmaybereproducedortransmittedinanyformorbyanymeans,electronic ormechanical,includingphotocopying,recording,oranyinformationstorageandretrievalsystem, withoutpermissioninwritingfromthepublisher.Detailsonhowtoseekpermission,further informationaboutthePublisher’spermissionspoliciesandourarrangementswithorganizationssuch astheCopyrightClearanceCenterandtheCopyrightLicensingAgency,canbefoundatourwebsite: www.elsevier.com/permissions. Thisbookandtheindividualcontributionscontainedinitareprotectedundercopyrightbythe Publisher(otherthanasmaybenotedherein). Notices Knowledgeandbestpracticeinthisfieldareconstantlychanging.Asnewresearchandexperience broadenourunderstanding,changesinresearchmethods,professionalpractices,ormedicaltreatment maybecomenecessary. Practitionersandresearchersmustalwaysrelyontheirownexperienceandknowledgeinevaluating andusinganyinformation,methods,compounds,orexperimentsdescribedherein.Inusingsuch informationormethodstheyshouldbemindfuloftheirownsafetyandthesafetyofothers,including partiesforwhomtheyhaveaprofessionalresponsibility. Tothefullestextentofthelaw,neitherthePublishernortheauthors,contributors,oreditors,assume anyliabilityforanyinjuryand/ordamagetopersonsorpropertyasamatterofproductsliability, negligenceorotherwise,orfromanyuseoroperationofanymethods,products,instructions,orideas containedinthematerialherein. LibraryofCongressCataloging-in-PublicationData AcatalogrecordforthisbookisavailablefromtheLibraryofCongress BritishLibraryCataloguing-in-PublicationData AcataloguerecordforthisbookisavailablefromtheBritishLibrary ISBN:978-0-323-98887-2 ISSN:1573-448X ForinformationonallNorth-Hollandpublications visitourwebsiteathttps://www.elsevier.com/books-and-journals Publisher:ZoeKruze AcquisitionsEditor:JasonMitchell EditorialProjectManager:NaizaErminMendoza ProductionProjectManager:DennyMansingh Designer:VictoriaPearson TypesetbyVTeX Contributors NikhilAgarwal MITDepartmentofEconomics,Cambridge,MA,UnitedStates JohnAsker UCLA,LosAngeles,CA,UnitedStates KevinA.Bryan UniversityofToronto,RotmanSchoolofManagement,Toronto,ON,Canada EricBudish UniversityofChicagoBoothSchoolofBusiness,Chicago,IL,UnitedStates RobertClark Queen’sUniversity,Kingston,ON,Canada LiranEinav DepartmentofEconomics,StanfordUniversity,Stanford,CA,UnitedStates AmyFinkelstein DepartmentofEconomics,MassachusettsInstituteofTechnology,Cambridge, MA,UnitedStates BenHandel UCBerkeley,Berkeley,CA,UnitedStates KateHo PrincetonUniversity,Princeton,NJ,UnitedStates AliHortaçsu UniversityofChicago,Chicago,IL,UnitedStates Jean-FrançoisHoude UniversityofWisconsin-Madison,Madison,WI,UnitedStates JakubKastl PrincetonUniversity,Princeton,NJ,UnitedStates RyanKellogg HarrisPublicPolicy,UniversityofChicago,Chicago,IL,UnitedStates NealeMahoney DepartmentofEconomics,StanfordUniversity,Stanford,CA,UnitedStates xi xii Contributors VolkerNocke UniversityofMannheim,Mannheim,Germany IsabellePerrigne RiceUniversity,Houston,TX,UnitedStates MarReguant DepartmentofEconomics,NorthwesternUniversity,Evanston,IL,UnitedStates HeidiL.Williams StanfordUniversity,Stanford,CA,UnitedStates Introduction to the series TheaimoftheHandbooksinEconomicsseriesistoproduceHandbooksforvarious branches of economics, each of which is a definitive source, reference, and teach- ingsupplementforusebyprofessionalresearchersandadvancedgraduatestudents. Each Handbook provides self-contained surveys of the current state of a branch of economicsintheformofchapterspreparedbyleadingspecialistsonvariousaspects ofthisbranchofeconomics.Thesesurveyssummarizenotonlyreceivedresultsbut also newer developments, from recent journal articles and discussion papers. Some originalmaterialisalsoincluded,butthemaingoalistoprovidecomprehensiveand accessiblesurveys.TheHandbooksareintendedtoprovidenotonlyusefulreference volumes for professional collections but also possible supplementary readings for advancedcoursesforgraduatestudentsineconomics. KennethJ.ArrowandMichaelD.Intriligator xiii Preface We are very happy to introduce you to the next two volumes of the Handbook of IndustrialOrganization.Asthelatestvolumeintheserieswaspublishedin2007,our contributorshavestriventodescribedevelopmentsinthefieldofindustrialorganiza- tioninthelastdecadeandahalf.Thishasbeenafruitfultimeforthefield,inwhich the“econometricrevolution”thatstartedwiththeNewEmpiricalIndustrialOrgani- zation(NEIO)approachdescribedinVolumes1and2ofthisHandbookcontinuedto gainmaturity,alongsideimportantdevelopmentsinthemicroeconomictheoryland- scape.Perhapsmoststrikingly,thetheory-basedempiricalmethodstoolboxthathas now become the mainstay of industrial organization economists has expanded its reach in a variety of application domains, covering broad swathes of the modern economy.Thisisnotsurprising,asthelifebloodofanyempiricalenterpriseisdata, anddatahasbecomequiteabundantthankstotheincreasinguseofinformationand computing technologies by modern economic enterprises. The data comes in many shapes and sizes, encompassing firm level data that powers productivity studies, to product and even consumer level data that allows researchers to estimate demand withahighleveloffidelity.Theabundanceofsuchdataopensupmanynewoppor- tunitiesaswellaschallenges,aswillbediscussedbythechaptersinthesenexttwo volumes. We have chosen to organize Volumes 4 and 5 in a thematic manner. The first 5 chapters, in Volume 4, review developments in the empirical/econometric method- ology domain, which is the continuationof the NEIO program. Chapters 6 through 9,alsoinVolume4,coverimportantdevelopmentsinindustrialorganizationtheory, muchofwhichhasbeenaccompaniedbyimportantempiricalapplicationsthathave taken these theories to data. Chapters 10 through 17, in Volume 5, are focused on applicationsofthetheoryandeconometrictoolboxtolongstandingeconomicpolicy questions and the in-depth study of important sectors of the economy. The policy questionsrangefromthedesignofmatchingandauctionmarkets,toregulation,an- titrust,andinnovationpolicies.Theexpandingreachofthe“IO”approachisfurther demonstratedinsectoralstudiesoncreditandinsurancemarkets,financialmarkets, healthcare,theenergysector,andtheenvironment.Theseareapplicationareaswhere therehasbeenanexplosioninmethodsandmodelsthatareofusenotjustforthose applications but also more broadly. That is, methods developed for application to these markets have been used elsewhere in Industrial Organization, as well as in these particular areas. This is one reason why they merit particular consideration here.Further,theresultsofmanyofthesepapershavehadmeasurableeffectsonthe realworld,forexampleonantitrustpolicyandenforcement. Chapter 1, titled “Foundations of Demand Estimation,” by Steven Berry and PhilipHaile,providesaunifieddiscussionofthekeyconceptualideasthathavebeen xv xvi Preface developedby several literatures on demandestimation.Maintaininga focus on dis- crete choice approaches popular in industrial organization applications, the chapter beginsbydiscussingtheuniqueanddifficultchallengesposedbytheproblemofde- mandestimation.Thisisfollowedbydiscussionsonthechoiceofinstrumentalvari- ables,theutilizationofdifferentdatatypesthataretypicallyavailable(market-level, consumer-level,andpaneldata),andthequestionofnonparametricidentification. Chapter 2, titled “Empirical Models of Demand and Supply in Differentiated Products Industries,” by Amit Gandhi and Aviv Nevo, complements Chapter 1 by focusingontheimplementationandapplicationsofmoderndemandestimationmeth- ods utilized by industrial organization researchers. The chapter opens with a brief recapoftheclassicpaperbyBresnahan(1987)thathassettheframeworkformany empiricalpapers:formulateandestimateamodeldifferentiatedproductdemand(in Bresnahan’s case, for automobiles), and use demand estimates to infer unobserved supply side parameters and test hypotheses about firm conduct. Gandhi and Nevo thendiscusssubsequentdevelopmentsinthedemandestimationliterature,especially discrete choice demand models, reviewing how different types of data and sources of exogenous variation/instrumental variables can be used to identify and estimate theparametersofinterest.Theythendiscussapplicationsofdemandestimates,such assimulatingtheeffectsofmergers,andtestingalternativemodelsoffirmconduct, withcloseattentiontotheverticalorganizationofthesupplyside.Thechapteralso discusses extensions to the static discrete choice demand model, including models allowing for multiple purchases, and dynamic considerations such as storable and durablegoods. Chapter3is“AnIndustrialOrganizationPerspectiveonProductivity,”byJanDe LoeckerandChadSyverson.Thechapteropenswithadiscussionoftheconceptof productivity and some motivating stylized empirical facts about firm and establish- mentlevelmeasuredproductivity:dispersionacrossproducers,persistencewithina producer,andthecorrelationofproductivitywithotherproducer-leveloutcomessuch asprofitability,size,survival,priceschargedtocustomers,andwagespaidtowork- ers.Thechapterthendiscussesanequilibriummodelthatcangeneratethesestylized facts.Therestofthechapterfocusesonempiricalworkonproductivity.Startingwith a discussionof themeasurementof outputsand inputs,it discusses the factor share andproductionfunctionapproachestoproductivitymeasurement,focusingonchal- lengesintroducedbytheendogeneityandmismeasurementofinputs,anddepartures fromassumptionsofperfectcompetition,homogeneousproducts,andsingleproduct firms/establishments.Itthendiscussesseveralsignificantapplicationsofproductivity analysis,rangingfromtheanalysisofproductivitydrivers/dispersiontothe(mis)al- locationofinputsacrossenterpriseswithdifferentlevelsofproductivity.Thechapter concludes with a “look ahead” to some recent developmentsin the productivity lit- erature, especially the study of product market and labor market power using the productionfunctionapproachthathasgainedconsiderableattentioninrecentyears. Chapter 4 is “Dynamic Games in Empirical Industrial Organization,” by Victor Aguirregabiria, Allan Collard-Wexler, and Stephen Ryan. The theory and econo- metricsofdynamicgamesarecentraltounderstandingtheendogenousevolutionof Preface xvii marketstructureandcompetitiveoutcomes.ThechapterlaysouttheMarkovPerfect Nash Equilibrium (MPNE) framework that has found many applications in indus- trialorganization.Itthenoverviewsthetypesofdatathataretypicallyavailablefor empirical research, and discusses whether and how model parameters can be iden- tified using available data. Estimation of structural parameters of MPNE models is discussed next, with a focus on addressing implementation challenges introduced by computational complexity and curse of dimensionality. The chapter concludes withadiscussionofthemanyempiricalapplicationsoftheMPNEframework.The applicationtopicsdiscussedincludeinnovation,antitrustandmergers,regulation,un- certainty and investment, advertising, dynamic pricing, product repositioning, and othertopics. Chapter 5 is “Moment Inequalities and Partial Identification in Industrial Or- ganization” by Brendan Kline, Ariel Pakes, and Elie Tamer. This chapter presents recent developments in the use of moment inequality models and partial identifica- tion more generally in Empirical Industrial Organization. It begins by formalizing thedistinctionsbetweenthedifferentnotionsofidentificationandtheirrelationship to various estimators. The following sections provide a set of assumptions that are useful in enabling the researcher to use revealed preference inequalities for estima- tion.Theresultingframeworkallowsforstrategicinteraction,alternativeinformation sets, and measurement and mis-specification error. A particular implementation is then covered in detail, motivated by the standard discrete choice literature. Alter- native assumptions are also considered that allow the researcher to analyze likely counterfactual equilibria. Estimation and inference are considered in detail, includ- ing the problems that arise in analyzing interacting agent models with a complete informationassumption.Overall,thechaptershouldbeusefultoappliedresearchers in I.O. who might wish to use these methods in their own research, helping them understandmanyoftheprimaryideasunderlyingtheiruseinthepreviousliterature. Chapter6is“FrictionsinProductMarkets”byAlessandroGavazzaandAlessan- dro Lizzeri. This chapter considers several types of frictions that impede efficient transactions.Thefocusisonmarketfrictionsratherthanfirm-levelfrictions,although the presence of such frictions may lead to the emergence of intermediaries or deal- ers.FollowingtherecentliteratureinIndustrialOrganization,thefocusislargelyon markets with large numbers of participants rather than on transactions with a small numberofplayers.Thesectionsofthechapterconsidertransactionscosts;theiref- fects in secondary markets for durable goods; asymmetric information; and search andmatchingfrictions.Ineachsection,theauthorsconsiderhowtherelevantfriction affectsallocationsandconsumerbehavior;discusstheeffectofthesefrictionsonin- centives;andalsoconsiderhowthepresenceoffrictionsgivesscopeforthirdparties suchasintermediariestoplayanimportantroleinsomemarkets. Chapter7,titled“Two-sidedMarkets,Pricing,andNetworkEffects,”isbyBruno Jullien,Alessandro Pavan,and Marc Rysman. This chapter overviews the literature ontheeconomicsofplatforms.Platformsareintermediariesbetweendifferentmarket sides.Examplesareride-hailingcompaniesthatmediatethedeliveryoftransportation servicesbydriversanditsdemandbyriders;creditcardcompaniesthatofferpayment xviii Preface servicesbetweenbuyersandsellers;cableTVcompaniesthatmediatebetweencon- tentprovidersandviewers;andmanyothers.Platformsoftendisplaynetworkeffects inthatthevalueofbeingontheplatformdependsonwhoelseisalsoontheplatform. Platforms pricingdecisions canhavecomplexconsequencesthat are quitedifferent fromstandardpricingintuitionsbecausetheeffectsofapricechangeononesideof themarkethasrepercussionsforthewillingnesstopayoftheotherside.Thechapter surveysboththerapidlygrowingtheoreticalliterature,andthemorerecentempirical contributions. Chapter8,titled“InformationMarketsandNonmarkets,”isbyDirkBergemann andMarcoOttaviani.Thischapterexaminesmarketfailuresandpotentialpoliciesin thecontextofinformationgoods,andalsoconsiderstheinteractionbetweenmarket andnon-marketincentives.Thetopicisparticularlytimelygiventheincreasingrele- vanceofdataformanytransactions,andtheimportanceofdataforsomeofthemost valuable(intermsofmarketcapitalization)companiesintheworld,suchasAmazon, Google,andFacebook.Thechapterfirstoverviewsrecentcontributionstothetopic ofmechanismdesignapproachestobuyingandsellinginformation;itthendiscusses the organization of markets for information; it then moves on to consider tools that facilitate or impede the trade of information; it next discusses institutional arrange- ments for the collection of information and for aggregating expert judgments such asforecasts.Finally,itdiscussestheliteratureonscienceasaformofproductionof informationand knowledgethat is not organized as market. These themes come up againinChapter13oninnovation. Chapter 9 is “Structural Empirical Analysis of Contracting in Vertical Markets” byRobinS.Lee,MichaelD.Whinston,andAliYurukoglu.Thepapersynthesizesthe developing literature on vertical markets in Industrial Organization over the last 15 years.Recentempiricalworkonverticalandintermediategoodsmarketshasutilized advances in the theory of vertical contracting and in empirical demand estimation literatures. This chapter pulls together the somewhat different theoretical structures from different papers into a single, useful framework. It is clear that assumptions regardingthenatureofthesupplychain,whereseverallayersmaybecharacterized byanoligopolisticmarketstructure,andregardinghowcontracttermsarenegotiated, are often important for the predicted effects of counterfactual policy choices. The chapteralsoconsiderstheavailabilityofdataindifferentverticalmarkets. Chapter10,titled“MarketDesign,”isbyNikhilAgarwalandEricBudish.This chapter surveys both theoretical and empirical contributions in the rapidly growing field of market design. As the authors frame it, the market design field studies the rulesgoverningmarketinteractionsamongvariousparticipantsinordertounderstand theireffects,toidentifypotentialmarketfailures,andtoremedythembydesigning betterinstitutions.Onepossiblealternativedefinitionofthefieldisas“appliedmech- anismdesign,”offeringdiagnosesofproblemsinexistinginstitutions,andsuggesting workableimprovements.Classicexamplesofimportantsettingswheremarketdesign hashadanimportantinfluenceisthecaseofschoolchoice,wherestudentsmustbe matchedtoschools,auctions,kidneyexchange,thedesignofsuitablefinancialmar- ket trading mechanisms, etc. The field has generated important theoretical insights Preface xix startingwithworkinthe1960s,andhasmorerecentlyseenaflourishingofempirical work;overalltheseinsightshavebeeninfluentialinaffectingpolicyinanumberof settings.Thechapterthoroughlydiscussesthesevariousangles. Chapter11,titled“EmpiricalPerspectiveson Auctions,”is byAli Hortaçsuand IsabellePerrigne.Auctionsareusedtoallocateawidevarietyofgoodsrangingfrom paintings,toelectricity,totreasurysecurities,toadvertisingslotsontheinternet.A vastliteraturehasemergedtostudythepropertiesofvariousauctionformats.Since thelate1980stheempiricalanalysisofauctionshasgrowntremendously.Thischap- ter provides an overview of the field which offers both an update and a different perspectivecomparedwiththechapterbyHendricksandPorterinthepreviousvol- umeoftheHandbookofIndustrialOrganization.HendricksandPortercombineall aspectsoftheanalysisofauctiondatabystressingthelinkbetweenauctiontheory, empirical practice and public policy questions. The present chapter instead is orga- nizedbythetypeofgoodsoldbecausethepolicyanddesignquestionsoftendepend onthenatureoftheproductauctioned.Incontrastwithpriorsurveysthisonefocuses onpolicyanddesignquestionsratherthaneconometricmethods. Chapter12,titled“Collusion,Mergers,andRelatedAntitrustIssues”,isbyJohn AskerandVolkerNocke.Thechaptersurveystherecentliteratureonantitrustregu- lation,consideringinparticularquestionsrelatingtocollusionandmergers.Interms of collusion, the authors begin by asking how pervasive collusive conduct is, and thenanalyzehowcollusiveschemesoperate.Theyalsoconsiderboththeoreticaland empirical work on the impact of collusion on market outcomes. They then turn to mergers. The discussion considers in particular the impact of enforcement institu- tions,workonunilateraleffects,andtheoryrelatingtotheselectionofwhichmergers areproposedtoantitrustagenciesandoptimalpolicyinthefaceofthatselection.The smallliteratureconsideringoptimalremediesisalsodiscussed,togetherwiththeem- piricalevidenceontheeffectsofmergers. Chapter13,titled“Innovation:MarketFailuresandPublicPolicies,”isbyKevin A. Bryan and Heidi L. Williams. Innovation is responsible for a large portion of economic growth, and understanding its drivers as well as evaluating policies that mayencourageinnovationisclearlyoffirstorderimportance.Economistshavelong understood that competitive markets, which have admirable efficiency properties in othersettings,haveimportantshortcomingsinthecontextofinnovation,andpublic policyhasrespondedbyintroducingpotentialotherdistortions,suchasinthecontext ofpatentpolicywhichgrantsmonopolypowertoaninventor.Thischapterdiscusses these theoretical and policy issues, then examines the many challenges facing the empiricalresearcheraimingtomeasuretheextentofinefficienciesandthepotential effectivenessofalternativepolicies.Thepaperalsodiscussestheinteractionbetween marketandnon-marketincentivesforscienceandinnovation. Chapter 14 is “The IO of Selection Markets” by Liran Einav, Amy Finkelstein, andNealeMahoney.ThevastmajorityoftheoreticalandempiricalresearchinIndus- trialOrganizationfocusesonmarketswheretheidentityofthebuyerdoesnotaffect thecost.Insuchmarkets,sellerscareabouthowconsumerdemandaffectshowmany units they sell, but they do not care which consumers buy and which do not. In se-