1、外文文献翻译译文原文EMPIRICALRESEARCHONACCOUNTINGCHOICE1EARNINGSMANAGEMENTBURGSTAHLERANDDICHEV1997REPORTTHATMANAGERSAPPARENTLYMANAGEEARNINGSTOAVOIDEARNINGSDECREASESANDLOSSESTHEYRELYONATRANSACTIONSCOSTTHEORYRATHERTHANEFFICIENTCONTRACTINGORMANAGERIALOPPORTUNISMTOEXPLAINTHEIRRESULTSTHATIS,THEYSUGGESTTERMSOFTRANS
2、ACTIONSWITHSTAKEHOLDERSAREMOREFAVORABLEFORFIRMSWITHHIGHERRATHERTHANLOWEREARNINGSSEEBOWEN,ETAL,1995FORFURTHERDISCUSSIONOFTHISPOINTANDALSOTHATINVESTORSARENOTFULLYRATIONALINASSESSINGTHEINFORMATIONCONTENTOFREPORTEDEARNINGS,CONSISTENTWITHPROSPECTTHEORYTHEABOVESTUDIESALLREPORTEVIDENCEOFEARNINGSMANAGEMENTV
3、IACHOICESOFACCOUNTINGMETHODSBUTNONEDOCUMENTANYASSOCIATEDPRICEREACTIONSTOTHESECHOICESINOTHERWORDS,THESESTUDIESDONOTEXPLOREWHETHERTHESEACCOUNTINGCHOICESHAVEECONOMICIMPLICATIONSBARTH,ETAL1999,ONTHEOTHERHAND,FINDTHATFIRMSWITHATIMESERIESOFINCREASINGEARNINGSHAVEHIGHERPRICEEARNINGSMULTIPLESAFTERCONTROLLING
4、FORRISKANDGROWTH,THANFIRMSWITHOUTANINCREASINGEARNINGSPATTERNTHISEVIDENCEISCONSISTENTWITHTHESUCCESSOFEARNINGSMANAGEMENTHOWEVER,BARTHETALDONOTEXPLICITLYTESTFOREARNINGSMANAGEMENTANDDONOTATTRIBUTETHEEARNINGSPATTERNASNECESSARILYDUETOEARNINGSMANAGEMENTDAVIS1990,INAPARTIALREPLICATIONANDEXTENSIONOFTHEHONG,E
5、TAL1978STUDYOFTHEPURCHASEANDPOOLINGCHOICE,FINDSTHATACQUIRINGFIRMSTHATUSETHEPURCHASEMETHODENJOYPOSITIVEABNORMALRETURNSOVERTHEPERIODEXTENDINGFROMBEFORETHEANNOUNCEMENTOFTHEBUSINESSCOMBINATIONTOAFTERITSCONSUMMATIONACQUIRINGFIRMSUSINGTHEPOOLINGMETHODENJOYONLYNORMALMARKETRETURNSHISRESULTSARECONSISTENTWITH
6、THOSEOFHONG,ETAL2MULTIPLEMOTIVATIONSINADDITIONTOTHEPROBLEMOFADDRESSINGMULTIPLEACCOUNTINGCHOICES,GENERALLYASREFLECTEDINACCRUALS,THEREISALSOTHEISSUEOFMULTIPLE,ANDPOTENTIALLYCONFLICTING,MOTIVATIONSFORTHEACCOUNTINGCHOICESMOSTOFTHEWORKDISCUSSEDINSECTION4FOCUSESONASINGLEMOTIVEFORACCOUNTINGCHOICEDECISIONSF
7、OREXAMPLE,THECOMPENSATIONLITERATUREFOCUSESONTHEQUESTIONOFWHETHERMANAGERSUSEACCOUNTINGDISCRETIONTOMAXIMIZETHEIRCOMPENSATIONIMPLICITLY,THERESULTSSUGGESTTHATMANAGERSACTIONSCOMEATTHEEXPENSEOFSHAREHOLDERSBUTIFTHISISSO,WHYDOCOMPENSATIONCONTRACTSALLOWDISCRETIONONEPLAUSIBLEANSWERISTHATMANAGERSACTIONSARENOTO
8、NLYANTICIPATED,BUTALSODESIRABLEFROMSHAREHOLDERSPERSPECTIVEFOREXAMPLE,THESAMEACCOUNTINGCHOICESTHATMAXIMIZEMANAGERSCOMPENSATIONMAYALSODECREASEBONDCOVENANTVIOLATIONSORINCREASEASSETVALUATIONSHOWEVER,SUCHMOTIVESARETYPICALLYNOTINCLUDEDINTHEANALYSISBYFOCUSINGONONEGOALATATIME,MUCHOFTHELITERATUREMISSESTHEMOR
9、EINTERESTINGQUESTIONOFTHEINTERACTIONSBETWEENANDTRADEOFFSAMONGGOALSMOREOVER,ITISNOTCLEARWHETHERTHECONCLUSIONSAREATTRIBUTABLETOTHESPECIFICMOTIVATIONBEINGANALYZEDGENERALLYRESULTSCONSISTENTWITHONEHYPOTHESISARECONSISTENTWITHMANYFOREXAMPLE,WHATMAYAPPEARTOBEANOPPORTUNISTICCHOICEOFANEARNINGSINCREASINGACCOUN
10、TINGMETHODCHOICETOBENEFITMANAGERSATTHEEXPENSEOFOTHERSTAKEHOLDERSINTHEFIRM,MAYBEINFACTARESPONSETOAVOIDABONDCOVENANTVIOLATIONANDTHUSBENEFITSALLOTHERSTAKEHOLDERSATTHEEXPENSEOFTHECREDITORSFINALLY,WITHONLYFEWEXCEPTIONS,RESEARCHINTHE1990SGENERALLYFOCUSESONMOTIVESIDENTIFIEDINTHE1970SAND1980STYPICALLYTHEUSU
11、ALSUSPECTSAREROUNDEDUPHOWEVER,WESUSPECTTHATNEWINSIGHTSMAYBEGAINEDBYINVESTIGATINGADDITIONALMOTIVESTHEPROBLEMOFMULTIPLECONFLICTSCANBEVIEWED,INTURN,ASASPECIALCASEOFTHEFAMILIAR“CORRELATEDOMITTEDVARIABLE”PROBLEMINECONOMETRICSFOREXAMPLE,IFFIRMSWITHCOMPENSATIONCONTRACTSTHATDEPENDONEARNINGSAREALSOMORELIKELY
12、TOHAVEHIGHPOLITICALCOSTS,THENACOMPENSATIONSTUDYCOULDFINDARELATIONSHIPBETWEENCOMPENSATIONCONTRACTSANDACCOUNTINGCHOICE,EVENIFTHEACCOUNTINGCHOICESAREINFACTDRIVENBYPOLITICALCONSIDERATIONSTHEUSUALSOLUTIONTOTHISPROBLEMISTOADDCONTROLVARIABLESHOWEVER,THISSOLUTIONASITISAPPLIEDTOACCOUNTINGCHOICERESEARCHSUFFER
13、SFROMATLEASTTHREEDRAWBACKSFIRST,RESEARCHERSOFTENRELYONCOARSEORINAPPROPRIATEPROXIESTOMEASURETHEROLEOFTHEOMITTEDDETERMINANTSOFACCOUNTINGCHOICEFOREXAMPLE,INCOMPENSATIONSTUDIES,LEVERAGEANDSIZEVARIABLESHAVEBEENUSEDTOPROXYFORBONDCOVENANTEFFECTSANDPOLITICALCOSTSSECOND,INFERENCEPROBLEMSARELIKELYTOARISEWHENA
14、NALYZINGMULTIPLEMOTIVATIONSUSINGPROXIESWITHDIFFERINGAMOUNTSOFMEASUREMENTERROR,PARTICULARLYWHENTHEUNDERLYINGEFFECTSARECORRELATEDINSUCHINSTANCES,THEPROXYWITHTHELEASTAMOUNTOFMEASUREMENTERRORISLIKELYTODOMINATE,EVENWHENITSTRUEEFFECTMAYNOTBETHEMOSTIMPORTANTFOREXAMPLE,ASSUMETHATPOLITICALCOSTSANDCOMPENSATIO
15、NMOTIVATIONSARECORRELATED,BUTTHATACCOUNTINGCHOICESAREDRIVENBYPOLITICALCOSTSNEVERTHELESS,IFSIZEISUSEDASANOISYPROXYFORPOLITICALCOSTSWHILETHERESEARCHERISABLETOMEASURETHEDESIGNOFTHECOMPENSATIONCONTRACTWITHGREATPRECISION,THENAREGRESSIONOFACCOUNTINGCHOICEONSIZEANDCOMPENSATIONCONTRACTDESIGNWILLLOADONTHEMOR
16、EPRECISEBUTLESSECONOMICALLYIMPORTANTCOMPENSATIONVARIABLE,RATHERTHANTHEECONOMICALLYSIGNIFICANTBUTNOISYPROXYFORPOLITICALCOSTSLITTLE,IFANY,PROGRESSHASBEENMADEINADDRESSINGTHESECONCERNSFINALLY,THEPROBLEMOFMULTIPLEMOTIVATIONSISFURTHERCOMPOUNDEDBYTHEABSENCEOFLINEARITYSPECIFICALLY,DISCUSSIONOFTHECORRELATEDO
17、MITTEDVARIABLEISSUEOFTENATLEASTIMPLICITLYASSUMESTHATTHEVARIABLESBOTHTHOSEINCLUDEDANDTHOSEOMITTEDARELINEARANDTHATTHEOMITTEDVARIABLEISADDITIVELYSEPARABLEFROMTHEVARIABLEOFINTERESTTHESEASSUMPTIONSAREMADETOFACILITATERESEARCHDESIGNHOWEVER,THEREISNOEVIDENCETHATSUCHASSUMPTIONSAREJUSTIFIEDEVENTHOSEPAPERSTHAT
18、CONSIDERMULTIPLEMOTIVATIONSGENERALLYTREATTHEMINDEPENDENTLYINPRACTICE,OFCOURSE,MANAGERSFACEAVARIETYOFCONFLICTS,WHICHWILLNOT,INGENERAL,SUGGESTCONSISTENTCOURSESOFACTIONSUNDERTHESECIRCUMSTANCES,FIRMSMUSTMAKETRADEOFFSAMONGTHEVARIOUSGOALS3CONCLUSIONANDSUGGESTIONSFORFUTUREWEDONOTWANTTOLEAVETHEIMPRESSIONTHA
19、TRESEARCHERSHAVEGAINEDNOKNOWLEDGEOFTHEROLEANDTHEIMPORTANCEOFACCOUNTINGCHOICERATHER,OURCONCERNISTHATPROGRESSHASSLOWEDINPART,THISISDUETOUNAMBITIOUSATTEMPTSTOEXPANDTHEFIELDFOREXAMPLE,TESTINGTHEIMPLICATIONSOFONEMOREACCOUNTINGSTANDARDADDSVERYLITTLETOTHECUMULATIVEKNOWLEDGEAMOREINTRANSIGENTPROBLEMISTHEDIFF
20、ICULTYINSPECIFYINGRESEARCHDESIGNSTHATACCOMMODATETHECOMPLEXITYOFTHETASKATHANDTHATIS,THESIMULTANEOUSIMPACTOFMULTIPLECHOICES,MULTIPLEGOALSANDECONOMETRICCOMPLICATIONSRATHERTHANCONTINUETOREPLICATEWELLKNOWNRESULTSINSLIGHTLYDIFFERENTSETTINGS,WEFEELTHATITISIMPORTANTFORRESEARCHERSTOGRAPPLEWITHTHESEMOREDIFFIC
21、ULT,ANDATHEART,MOREFUNDAMENTALISSUESWEHAVETHREESPECIFICRECOMMENDATIONSFORFUTURERESEARCHFIRST,RESEARCHRESULTSFAILTOPROVIDECOMPELLINGEVIDENCEOFTHEIMPLICATIONSOFALTERNATIVEACCOUNTINGMETHODSANDWERECOMMENDMOREEFFORTSTODETERMINETHENATUREOFSUCHIMPLICATIONSTHELITERATUREPROVIDESAMPLECIRCUMSTANTIALEVIDENCETHA
22、TACCOUNTINGCHOICEMATTERSBUTLITTLEDIRECTEVIDENCEFOREXAMPLE,THELITERATUREDOCUMENTSTHATMANAGERSMAKEACCOUNTINGCHOICESCONSISTENTWITHBONUSMAXIMIZATION,BUTDOESNOTDETERMINEWHETHERTHISBEHAVIORRESULTSININCREASEDCASHPAYOUTSEVENIFSUCHEVIDENCEWEREPRODUCED,THENEXTSTEPWOULDBETOASCERTAINWHETHEREXPECTEDTOTALCOMPENSA
23、TIONINCREASESANDWHETHERTHISRESULTWASINTENDEDAND/ORANTICIPATEDBYTHECONTRACTINGPARTIESTHATIS,IFACCOUNTINGCHOICEHASSUCHPOTENTIALWEALTHIMPLICATIONS,THENCONTRACTINGPARTIESSHOULDPRICEACCOUNTINGCHOICEONEMODESTATTEMPTATTHISISSUEWASMADEBYHEALY,KANG,ANDPALEPU1987FORTHECASEOFMANAGERIALCOMPENSATION,BUTMOREISNEE
24、DEDANOTHERAVENUEISTOINVESTIGATETHECOSTSCOMPANIESAREWILLINGTOINCURTOMAINTAINACCOUNTINGMETHODCHOICEDISCRETIONAFIRSTATTEMPTATTHISAPPROACHCANBEFOUNDINBEATTY,ETAL2000,ANDWEENCOURAGEMOREEXPLORATIONOFTHISISSUESIMILARLY,THEREISNOCONSISTENTEVIDENCESUPPORTINGCLAIMEDVALUATIONDIFFERENCESDUETOACCOUNTINGMETHODSWE
25、DONOTKNOWWHETHERTHISISBECAUSEACCOUNTINGDIFFERENCESDONOTAFFECTFIRMVALUATIONORBECAUSEEMPIRICALMETHODSAREINADEQUATETODETECTANYSUCHEFFECTALTHOUGHMUCHEXISTINGRESEARCHISTIEDTOSTOCKRETURNS,ACCOUNTINGNUMBERSGENERALLYEXPLAINONLYASMALLPORTIONOFTHEVARIABILITYINSTOCKRETURNS,RAISINGTHEQUESTIONOFWHETHERTHISISTHEA
26、PPROPRIATEPLACETOLOOKFORANEFFECTDESPITEEXTENSIVEACADEMICRESEARCH,THEPROCESSBYWHICHSECURITYPRICESARESET,INCLUDINGTHEINFLUENCEOFACCOUNTINGDATA,REMAINSUNKNOWNSECOND,BECAUSEACCOUNTINGISUSEDFORMANYPURPOSES,WEHAVEARGUEDTHATITISINAPPROPRIATETOANALYZEONEACCOUNTINGISSUEOREVENONEGOALINISOLATIONIDEALLY,ONEWOUL
27、DHAVEACOMPREHENSIVETHEORYOFACCOUNTINGCHOICE,BUTSUCHATHEORYISPRESENTLYNOTAVAILABLE,ANDITSDEVELOPMENTDOESNOTAPPEARIMMINENTBECAUSEOFTHECOMPLEXITIESINHERENTINSUCHAMODELANALYTICALMODELSCANHELPPROVIDEGUIDANCETORESEARCHERSINSTRUCTURINGTHEEMPIRICALEXPERIMENTS,INIDENTIFYINGAPPROPRIATEVARIABLES,ANDINFORMULATI
28、NGALTERNATIVEHYPOTHESESATPRESENT,MOSTANALYTICALMODELSARESOABSTRACTASTOOFFERONLYLIMITEDGUIDANCETOEMPIRICISTSWEDONOTWANTTOGIVETHEIMPRESSIONTHATWEUNDERESTIMATETHECOMPLEXITYOFTHISTASKFURTHERMORE,WEARECAUTIOUSLYOPTIMISTICTHATPROGRESSTOWARDTHENECESSARYTHEORETICALIMPROVEMENTSCANBEMADEABSENTSUCHACOMPREHENSI
29、VETHEORY,PROGRESSCOULDSTILLBEMADEIFRESEARCHERSWOULDEXPANDTHEFOCUSWITHINACATEGORYASDESCRIBEDINSECTION3THUS,RATHERTHANNARROWLYANALYZINGTHEIMPLICATIONSOFACCOUNTINGCHOICEONBONDCOVENANTS,RESEARCHERSSHOULDEXPANDANDANALYZETHEIMPLICATIONSFORINTERNALCONTRACTINGFOREXAMPLE,HOWDOTHEFEATURESOFEXISTINGBONDCOVENAN
30、TSINFLUENCETHESTRUCTUREOFTHEINCENTIVECOMPENSATIONCONTRACTWHATCANWEINFERABOUTTHEEXPECTATIONSOFTHEBOARDOFDIRECTORSBYEXAMININGSUCHRELATIONSINGENERAL,WEBELIEVETHATANALYSISWITHINACATEGORYISJUSTIFIEDBECAUSETHECOMMONALITIESOFISSUESAREMUCHMOREPRONOUNCEDWITHINACATEGORYTHANACROSSCATEGORIESTHIRD,TOMAKEFURTHERP
31、ROGRESSINPROVIDINGMORECOMPELLINGTESTSOFACCOUNTINGCHOICE,RESEARCHERSSHOULDDEVELOPMOREPOWERFULSTATISTICALTECHNIQUESANDIMPROVERESEARCHDESIGNSTHELITERATUREHASBEGUNTHISPROCESSBYEXAMININGTHEADEQUACYOFEXISTINGSTATISTICALMETHODSSUCHEFFORTSSHOULDBEEXPANDEDWITHTESTINGOFMOREALTERNATIVEMODELSWEDONOTWANTTOSUGGES
32、TTHATRESEARCHMUSTADDRESSEVERYCONCERNRAISEDINTHISSURVEYINORDERTOBECONSIDEREDSUCCESSFULWERECOGNIZETHATMANYOFTHEPROBLEMSARECOMPLEXANDPOSEDIFFICULTRESEARCHDESIGNISSUESWEDOFEEL,HOWEVER,THATTHEFIELDHASBECOMETOOCONSERVATIVEWITHTOOMANYRESEARCHERSCONTENTTOJUSTIFYAMETHODOLOGYBECAUSEOTHERSHAVEUSEDITGREATEREFFO
33、RTSTOEMPLOYNEWMETHODOLOGIESANDMOREACCEPTANCEOFSUCHMETHODOLOGIESCOULDADVANCETHEFIELDRECENTWORKBYHUNT,ETAL1996,BEATTY,ETAL1995,ANDKANGANDSIVARAMAKRISHNAN1995PROVIDESGOODEXAMPLESOFEXTENDINGTHEMETHODOLOGICALBOUNDARIESWITHTHEAPPLICATIONOFSIMULTANEOUSEQUATIONSANDINSTRUMENTALVARIABLESTECHNIQUESTOACCOUNTING
34、PROBLEMSNEWORATLEASTUNTESTEDBYACCOUNTINGRESEARCHERSMETHODOLOGIESSHOULDBEEXPLOREDIFWEARETOMOVEFORWARDTHISMEANSTHATACCOUNTINGRESEARCHERSMUSTSTAYABREASTOFNEWDEVELOPMENTSANDNEWAPPLICATIONSINRESEARCHDESIGNINADDITION,WEALSOBELIEVETHATRESEARCHERSSHOULDMAKEBETTERUSEOFTHEIREXPERTISEASACCOUNTANTSBOTHSMALLSAMP
35、LESTUDIESANDFIELDSTUDIESWOULDFITINTOTHISAPPROACHALTHOUGHTHESMALLERSAMPLESIZESRAISEISSUESOFGENERALIZABILITY,WEFEELTHATTHISAPPROACHWOULDCOMPLEMENTEXISTINGLARGESAMPLESTUDIESANDPROVIDEGREATERINSIGHTINTOTHEUNDERLYINGCAUSESOFTHEEMPIRICALLYOBSERVEDEFFECTSLARGESAMPLESTUDIESCONTINUETOPLAYANIMPORTANTROLEBECAU
36、SESMALLSAMPLESTUDIESEXACERBATETHEPROBLEMOFDETERMININGWHETHERTHERESULTSAREDUETOUNUSUALORPATHOLOGICALCASESRATHERTHANTOTHEGENERALUSEOFACCOUNTINGIN“NORMAL”DAYTODAYCIRCUMSTANCESTHISPROBLEMISALSOAGGRAVATEDBYAPUBLICATIONBIASDRIVENBYTHEFACTTHATPAPERSWITHOUTRESULTSAREGENERALLYLESSLIKELYTOBEPUBLISHEDFURTHERMO
37、RE,PUBLISHEDRESEARCHMAYNOTTESTTHEEXANTEHYPOTHESISTHATIS,AUTHORSMAYVARYRESEARCHDESIGNANDVARIABLEDEFINITIONSUNTILSIGNIFICANTRESULTSAREFOUNDCHRISTIE,1990THEIMPLICATIONSOFTHISBIASARENOTCLEARTHEREMAYBEEVENFEWERCASESOFSTATISTICALLYSIGNIFICANTRESULTSONACCOUNTINGCHOICESTHANISAPPARENTFROMTHEPUBLICATIONRECORD
38、SPERHAPSALLTHATISBEINGDOCUMENTEDISNOISEFUNDAMENTALLY,WEBELIEVEITISNECESSARYTOSTEPBACKFROMTHECURRENTRESEARCHAGENDA,ANDTODEVELOPTHE“INFRASTRUCTURE”SURROUNDINGTHEFIELDINASENSE,THEACCOUNTINGCHOICEFIELDHASBEENAVICTIMOFITSOWNPERCEIVEDSUCCESS,ANDHASOUTRUNTHEDEVELOPMENTOFTHEORIES,STATISTICALTECHNIQUESANDRES
39、EARCHDESIGNTHATARENECESSARYTOSUPPORTITWETHEREFOREARECALLINGFORARETURNTOWORKINTHESEBASICAREAS,BEFORETHEFIELDISABLETOADVANCEFURTHERACADEMICACCOUNTINGRESEARCHMUSTULTIMATELYADDRESSTHEFUNDAMENTALQUESTIONSOFWHETHER,UNDERWHATCIRCUMSTANCES,ANDHOWACCOUNTINGCHOICEMATTERSTHESEQUESTIONSAREDIFFICULTBECAUSEOFTHEC
40、OMPLEXITYOFTHEENVIRONMENTINWHICHACCOUNTINGCHOICESAREMADETHEREMAYBEMANYDIFFICULTTOOBSERVEANDMEASUREEFFECTSANDMOTIVATIONSSURROUNDINGEACHCHOICEWHILEWEHAVEUNDOUBTEDLYMADESOMEPROGRESSONTHESEQUESTIONS,MOSTOFTHATPROGRESSOCCURREDINTHE1970SAND1980SWEHOPETHATOURREVIEWWILLPROVIDEANIMPETUSTOMOREVIGOROUSPURSUITO
41、FTHESEFUNDAMENTALQUESTIONSSOURCETHOMASDFIELDSTHOMASZLYSLINDAVINCENTEMPIRICALRESEARCHONACCOUNTINGCHOICEEB/OLHTTP/PAPERSSSRNCOM/SOL3/PAPERSCFMABSTRACT_ID258519HTML,20010101/20011004译文会计选择的实证研究1盈余管理BURGSTAHLER和DICHEV（1997年）的报告显示，管理人员显然是通过盈余管理来避免和减少损失的。他们依靠交易成本理论而不是有效承包经营或管理机会主义来解释他们的结果。也就是说，他们建议与利益相关者的
42、交易的条件应该是更有利于企业有较高收入而不是较低收入，并且投资者没有完全合理的评估与前景理论相一致的收入报告的信息内容。上述研究报告证明了，应通过盈余管理选择会计方法，但没有任何记录反应出选择对价格的影响。换句话说，这些研究没有探讨这些会计选择是否具有经济意义。BARTH等（1999年），从另一方面，发现那些已控制风险和市盈率增长的企业在经历了一段时间的收入增长后，比那些没有盈利模式的企业获得成倍的价值增长收益。这一证据与成功的盈余管理相一致。但是BARTH等人没有对盈余管理做出明确的测试，并且没有必要的将盈利模式归因于盈余管理。DAVIS1990部分复制和扩展了HONG等人1978关于购买和
43、联合经营的选择的研究，发现那些使用购买法的并购企业，通过延长宣布企业合并到其完善这段期间，享有绝对的超常回报。而使用权益结合法的并购企业只享有正常的市场回报。其结果与HONG等人是一致的。2多重动机除了普遍反映在权责发生制的多个会计选择问题，还包括潜在冲突和动机方面的会计选择。在第4节所讨论的大部分工作集中于一个单一动机的会计选择决策。例如，报酬部分侧重于管理者是否利用会计酌情权最大限度地实现报酬最大化。这个结果隐含的表明了管理者的行为是以股东为代价的。但是，如果是这样，为什么报酬契约允许酌情权一个似乎合理的答案是，经理人的行为不仅是可以被预料到的，而且从股东角度出发也是可取的。例如，针对相同
44、的会计选择，既可以最大限度地满足管理人员的补偿，也可以减少债券公约侵犯或增加资产估值。但是，这种动机通常不包括在分析中。通过在同一时间专注于一个目标，多数文献错过了很多目标之间存在的更加有意义的权衡互动问题。但是，这些结论是否是基于具体动机分析得出的，尚且不清楚。通常结果与一个假设一致，便与很多假设一致。例如，会计方法应该选择一个似乎是可以增加收入机会的方法（经理人以公司中其他利益相关者的利益为代价收益），事实上也可能是为了避免违反债券公约（因此导致其他利益相关者的收益应由债权人负责）。最后，只有少数例外，对20世纪90年代的研究一般是基于70年代和80年代就出现的动机由普遍的推测聚拢得来的。
45、然而，我们推测新的见解可以通过研究额外的动机获得。相反，大量冲突问题可以被视为经济学中类似“相关遗漏变量”问题的特殊情况。例如，如果公司基于收益签订定索赔合同，将更容易产生较高的政治成本，那么从报酬方面研究，我们很容易发现，报酬契约与会计选择之间存在着一定的关系，即使事实上会计选择是受制于政治方面的考虑的。通常解决这类问题的方法就是加强变量控制，但是这种方法在被运用到会计政策选择时，至少注意避免三个问题的存在。第一，研究者一般是依据粗略的或不恰当的标准来衡量会计选择中被忽略的决定性因素。比如，在研究报酬问题时，将杠杆作用和变数的大小作为衡量契约合同效用和政治成本的标准指标。第二，推理问题可能在
46、使用标准指标分析多个动机时，出现一些不同的测量误差，尤其与潜在效果相关联时。在这种情况下，用有更少测量误差的标准可能就占有一定优势，即使它的真正作用也变的不那么重要了。例如，假设政治成本与报酬动机相关，但会计选择是出于政治成本的。不过，如果变数的大小被作为标准用于政治成本，那么研究人员就能够精确的衡量出报酬契约的企图，从而，会计选择的尺度和报酬契约动机的回归将基于更精确的报酬变量，而不是经济意义上的政治成本标准。在处理这些问题上几乎没有任何进展。第三，多重动机问题中还存在线性缺失的现象。具体而言，人们在讨论相关遗漏变量时，往往假定其是线性的，并且可以从变项中加总可分。这些假设本身就是为了便于研
47、究设计，即使那些研究认为多重动机是独立于假设的，也没有任何证据证明这些假设是合理的。当然，在实践中管理人员所面临的通常是不能一致接受的各种分歧。在这种情况下，公司必须在不同的目标中权衡取舍。3结论及对未来的建议我们不想给别人留下研究人员已经不能再获得关于会计选择的作用和重要性相关知识的印象，而宁愿研究虽然缓慢但进行着。在某种程度上，是由于想要虚心试图扩展这片领域。例如，在测试一个会计标准的多种隐含意义的时候，只运用到很少的控制性知识。还有一个更棘手的问题是难以将指定的研究设计适应手头任务的复杂性那就是，多个选择多个目标和计量同时存在时的影响，对研究人员来说不仅是在稍微不同的环境中重复已知的结果
48、，我认为，重要的是努力克服困难和本质上更加基础的问题。我在这里对未来的研究提出三点具体的建议。首先，研究结果并没有对可供选择会计方法的意义提出令人信服的证据，我建议应致力于明确这些意义的本质。大多数的文献只为会计选择问题提供充分的间接证据，而提供了较少的直接证据。例如，管理者在选择会计政策时，是遵循红利最大化而没有明确这种行为是否会增加费用的支出。即使提供了这种证据，下一步应该做的是查明预期的红利总额是否增加，这种结果是否会被立约者预先考虑到。也就是说，如果会计政策选择具有大量的潜在意义，那么立约者将会查明会计政策选择下的价格。HEALY,KANG,ANDPALEPU1987在这方面为企业经营
49、者报酬做出了尝试性研究，但还需要更多的研究。另一个途径，就是通过调查企业为维持会计方法选择的决定所愿意承担的费用支出。第一个尝试研究这一领域问题的是BEATTY2000等人，我们鼓励在这方面有更多的探索。同样的，目前还没有一致的证据证明会计方法不同会导致对评估的要求不用。我们不知道这是不是因为不同会计政策对企业评估没有影响，还是因为经验主义的方法不足以察觉出其任何作用。尽管许多现有的研究都围着股票回报，会计数字一般只能解释很小一部分股票回报的变化性，这就提出了这是否是找出效果的合适的方法。其次，由于会计可以用于许多用途，我认为只分析一个会计问题或一个孤立的目标是不恰当的。理想的情况下，人们应该使用一套全面的会计选择理论，但这种理论目前是不可用的，在这种模式下因为其固有的内在复杂性，它不会立即发展下去。分析模型则可以为研究人员在构建实证研究时提供帮助，能够确定适当的变量并制定替代假说。目前，大多数分析模型只是抽象的为经验主义提供有限指导。我不希望我们给别人的印象是低估了这项工作的复杂性，此外，我们审慎乐观的认为必要的理论能够改善进展。在缺少这样一个全面的理论的情况下，如果研究人员如第三部分所述的将类别的焦点扩大，仍然可以取得进展。因此，不是狭隘的分析有关债务契约对会计选择的影响，研究人员应该扩大和分析其对缔约的影响。比如，现有的债券契约的特征对激励报酬契约结构的影响。
