An empirical evaluation of accounting income numbe

AccountingResearchCenter,BoothSchoolofBusiness,UniversityofChicagoAnEmpiricalEvaluationofAccountingIncomeNumbersAuthor(s):RayBallandPhilipBrownReviewedwork(s):Source:JournalofAccountingResearch,Vol.6,No.2(Autumn,1968),pp.159-178Publishedby:BlackwellPublishingonbehalfofAccountingResearchCenter,BoothSchoolofBusiness,UniversityofChicagoStableURL::31/07/201208:04YouruseoftheJSTORarchiveindicatesyouracceptanceoftheTerms&ConditionsofUse,[e-mail]..BlackwellPublishingandAccountingResearchCenter,BoothSchoolofBusiness,UniversityofChicagoarecollaboratingwithJSTORtodigitize,preserveandextendaccesstoJournalofAccountingResearch.*andPHILIPBROWNtAccountingtheoristshavegenerallyevaluatedtheusefulnessofaccount-ingpracticesbytheextentoftheiragreementwithaparticularanalyticmodel.Themodelmayconsistofonlyafewassertionsoritmaybearigorouslydevelopedargument.Ineachcase,themethodofevaluationhasbeentocompareexistingpracticeswiththemorepreferablepracticesim-pliedbythemodelorwithsomestandardwhichthemodelimpliesallpracticesshouldpossess.Theshortcomingofthismethodisthatitignoresasignificantsourceofknowledgeoftheworld,namely,theextenttowhichthepredictionsofthemodelconformtoobservedbehavior.Itisnotenoughtodefendananalyticalinquiryonthebasisthatitsassumptionsareempiricallysupportable,forhowisonetoknowthatatheoryembracesalloftherelevantsupportableassumptions?Andhowdoesoneexplainthepredictivepowersofpropositionswhicharebasedonun-verifiableassumptionssuchasthemaximizationofutilityfunctions?Further,howisonetoresolvedifferencesbetweenpropositionswhicharisefromconsideringdifferentaspectsoftheworld?Thelimitationsofacompletelyanalyticalapproachtousefulnessareil-lustratedbytheargumentthatincomenumberscannotbedefinedsub-stantively,thattheylackmeaningandarethereforeofdoubtfulutility.'Theargumentstemsinpartfromthepatchworkdevelopmentofaccount-*UniversityofChicago.tUniversityofWesternAustralia.TheauthorsareindebtedtotheparticipantsintheWorkshopinAccountingResearchattheUniver-sityofChicago,ProfessorMyronScholes,andMessrs.OwenHewettandIanWatts.1VersionsofthisparticularargumentappearinCanning(1929);Gilman(1939);PatonandLittleton(1940);Vatter(1947),Ch.2;EdwardsandBell(1961),Ch.1;Chambers(1964),pp.267-68;Chambers(1966),pp.4and102;Lim(1966),esp.pp.645and649;Chambers(1967),pp.745-55;Ijiri(1967),Ch.6,esp.pp.120-31;andSterling(1967),p.65.159160JOURNALOFACCOUNTINGRESEARCH,AUTUMN,1968ingpracticestomeetnewsituationsastheyarise.Accountantshavehadtodealwithconsolidations,leases,mergers,researchanddevelopment,price-levelchanges,andtaxationcharges,tonamejustafewproblemareas.Becauseaccountinglacksanall-embracingtheoreticalframework,dissimi-laritiesinpracticeshaveevolved.Asaconsequence,netincomeisanag-gregateofcomponentswhicharenothomogeneous.Itisthusallegedtobeameaninglessfigure,notunlikethedifferencebetweentwenty-seventablesandeightchairs.Underthisview,netincomecanbedefinedonlyastheresultoftheapplicationofasetofprocedures{X1,X2,...}toasetofevents{Y1,Y2,-..}withnootherdefinitivesubstantivemeaningatall.Canningobserves:Whatissetoutasameasureofnetincomecanneverbesupposedtobeafactinanysenseatallexceptthatitisthefigurethatresultswhentheaccountanthasfinishedapplyingtheprocedureswhichheadopts.2Thevalueofanalyticalattemptstodevelopmeasurementscapableofdefinitiveinterpretationisnotatissue.Whatisatissueisthefactthatananalyticalmodeldoesnotitselfassessthesignificanceofdeparturesfromitsimpliedmeasurements.Henceitisdangeroustoconclude,intheabsenceoffurtherempiricaltesting,tha~talackofsubstantivemeaningimpliesalackofutility.Anempiricalevaluationofaccountingincomenumbersrequiresagree-mentastowhatreal-worldoutcomeconstitutesanappropriatetestofuse-fulness.Becausenetincomeisanumberofparticularinteresttoinvestors,theoutcomeweuseasapredictivecriterionistheinvestmentdecisionasitisreflectedinsecurityprices.3Boththecontentandthetimingofexistingannualnetincomenumberswillbeevaluatedsinceusefulnesscouldbeim-pairedbydeficienciesineither.AnEmpiricalTestRecentdevelopmentsincapitaltheoryprovidejustificationforselectingthebehaviorofsecuritypricesasanoperationaltestofusefulness.Anim-pressivebodyoftheorysupportsthepropositionthatcapitalmarketsarebothefficientandunbiasedinthatifinformationisusefulinformingcapitalassetprices,thenthemarketwilladjustassetpricestothatinformationquicklyandwithoutleavinganyopportunityforfurtherabnormalgain.4If,astheevidenceindicates,securitypricesdoinfactadjustrapidlytonewinformationasitbecomesavailable,thenchangesinsecuritypriceswillre-2Canning(1929),p.98.8AnotherapproachpursuedbyBeaver(1968)istousetheinvestmentdecisi