requirementscncn
cncn com 时间:2021-03-02 阅读:()
ThisPDFisaselectionfromapublishedvolumefromtheNationalBureauofEconomicResearchVolumeTitle:TaxPolicyandtheEconomy,Volume17VolumeAuthor/Editor:JamesM.
Poterba,editorVolumePublisher:MITPressVolumeISBN:0-262-16220-2VolumeURL:http://www.
nber.
org/books/pote03-1ConferenceDate:October8,2002PublicationDate:January2003Title:TheDivergencebetweenBookIncomeandTaxIncomeAuthor:MihirA.
DesaiURL:http://www.
nber.
org/chapters/c11538THEDIVERGENCEBETWEENBOOKINCOMEANDTAXINCOMEMihirA.
DesaiHarvardBusinessSchoolandNBEREXECUTIVESUMMARYThispaperexaminestheevolutionofthecorporateprofitbaseandtherelationshipbetweenbookincomeandtaxincomeforU.
S.
corporationsoverthelasttwodecades.
Thepaperdemonstratesthatthisrelationshiphasbrokendownoverthe1990s,andithasbrokendowninamannerconsistentwithincreasedtax-shelteringactivity.
Thepapertracesthegrowingdiscrepancybetweenbookincomeandtaxincomeassociatedwithdifferentialtreatmentsofdepreciation,thereportingofforeignsourceincome,andinparticularthechangingnatureofemployeecom-pensation.
Forthelargestpubliccompanies,proceedsfromoptionexer-cisesequaled27percentofoperatingcashflowfrom1996to2000.
Thesedeductionsappeartobefullyutilized,therebycreatingthelargestdistinc-tionbetweenbookincomeandtaxincome.
Whilethedifferentialtreat-mentoftheseitemshashistoricallyaccountedfullyforthediscrepancybetweenbookincomeandtaxincome,thispaperdemonstratesthatbookandtaxincomehavedivergedmarkedlyforreasonsnotassociatedwiththeseitemsduringthelate1990s.
In1998,morethanhalfthedifferenceIamgratefultoAlanAuerbach,JamesHines,PeterMerrill,LilMills,GeorgePlesko,JamesPoterba,andparticipantsattheInternationalSeminarinPublicEconomicsConferenceandtheTaxPolicyandtheEconomyConferenceforhelpfulcomments.
ResearchassistancefromMarkVeblen,JamesZeitler,andYumingZou,andfinancialsupportfromtheDivisionofResearchatHarvardBusinessSchoolaregratefullyacknowledged.
270Desaibetweentaxandbookincome—approximately$154.
4billion,or33.
7per-centoftaxincome—cannotbeaccountedforbythesefactors.
Thispaperproceedstodevelopandtestamodelofcostlytaxshelteringanddemon-stratesthatthebreakdownintherelationshipbetweentaxincomeandbookincomeisconsistentwithincreasinglevelsofshelteringduringthelate1990s.
Thesetestsalsoexploreanalternativeexplanationofthesere-sults—coincidentincreasedlevelsofearningsmanagement—andfindthatthenatureofthebreakdownbetweenbookandtaxincomecannotbeexplainedfullybythisalternativeexplanation.
1.
INTRODUCTIONRecenttrendsinthecorporateprofitbase,andthetaxrevenuesitgener-ates,havedrawnconsiderableattention.
Inparticular,thegapbetweenbookincomeandtaxincomehasreportedlywidenedoverthelastdecade,andcorporatetaxreceiptshavebeenlowerthanexpected.
Seenagainstthebackdropofaprotractedeconomicexpansionoverthe1990s,thesetrendshavegeneratedconsiderableconcernovertheproliferationoftax-shelteringactivitybycorporations.
Theseconcernshavebeenamplifiedfurtherbyanecdotalevidenceontheproliferationoftheseschemesandbyhigh-profilecasesassociatedwithlargecorporations.
Estimatesoftax-shelteringactivityhaverangedwidely,dependingonthesourceandmethodologyemployedingeneratingthoseestimates.
1Thefocusoncorporatetaxsheltersraisesseveralnewquestionsrelatedtotheintegrityofthecorporatetaxbase.
Inparticular,financialinnova-tionsthatreducethecostsof,andwidenthescopefor,recharacterizingincomemayleadtodifficultiesinmaintainingthecorporatetaxbase.
Sim-ilarly,theincreasedimportanceoflegitimateforeignoperationsforU.
S.
firms,theattendanttransfer-pricingopportunitiesaffordedbysuchoper-ations,andincreasedcompetitionbyjurisdictionsforthoseinvestmentsmayincreasethescopeforincomeshifting.
Inshort,U.
S.
firmsarefacedwithenhancedopportunitiesforavoidingorevadingcorporatetaxesthroughcheaper,moresophisticated,andlesstransparentmechanisms.
Thesequestionsandconcernsovertheviabilityofthecorporatetaxbaseandthescopeofshelteringactivityhavegivenrisetovariedlegisla-tiveproposalsbutlimitedempiricalanalysisoftheclaimsassociatedwithincreasedtax-shelteringactivity.
Additionally,evidenceonshelteringac-1Theconcernsregardingthescopeandincreasedincidenceoftaxshelterscanbefoundinvarioussources,includingBankman(1999),Sullivan(1999a,1999b,2000a)andU.
S.
Treasury(1999).
Talisman(1999)andKies(1999a,1999b,1999c,and2000)providealternativeinterpre-tationsofthesetrends.
Foraperspectiveonthevarietyofinstrumentsassociatedwithshel-teringactivity,seeU.
S.
Treasury(1999).
TheDivergenceBetweenBookIncomeandTaxIncome171tivitythatcomesfromreportedbookincomeandtaxincomeraisestheissueofotherdifferencesthatmightarisebetweenbookandtaxincomeandtheirgrowingincidenceormagnitude.
Thus,empiricalexplorationsofthenatureofshelteringactivitymustjointlyinvestigatethechangingnatureofbookandtaxincometoidentifyanyunderlyingtrendsinthegapsthatseparatethem.
2Thispaperattemptstoilluminaterecenttrendsinthecorporateprofitbasetohighlightthepossibleaggregatescopeofshelteringactivity,theotherdeterminantsofthegapsbetweenbookandtaxincomeandtheirscopeforexplainingrecenttrends,andthedynamicsoftherelationshipbetweentaxandbookincomeoverthelastdecade.
Theevidenceprovidedinthispaperdemonstratesthatthelinkbetweenbookandtaxincomehasbrokendownoverthelastdecadefortworeasons.
First,theidentifi-ablefactorsthathavebeenassociatedtraditionallywiththedistinctionbetweenbookandtaxincomehavegrowntremendously.
Inadditiontothegrowingrelevanceofoverseasoperationsandthedifferentmeasuresofdepreciation,thepaperexplorestheproliferationofemployeestockoptions(ESOs)andquantifiestheirimpactonthecorporateprofitbase.
From1996to2000,netproceedsfromtheexerciseofESOsforthelargestU.
S.
corporationsapproximated27percentofoperatingcashflows.
Thesenetproceedsappeartobeutilizednearlyfullyasdeductionsagainstpre-taxincomeforthecorporationsstudied.
Second,thebreakdownbetweentaxincomeandbookincomeisnotlimitedtothegrowthofthesetraditionaldistinctionsbetweentaxandbookincome.
Whilethedistinctivetreatmentoftheseitemshashistori-callyaccountedfullyforthedifferencebetweenbookandtaxincome,thepaperdemonstratesthatbookandtaxincomehavedivergedmarkedlyforreasonsnotassociatedwiththeseitemsduringthelate1990s.
In1998,morethanhalfofthedifferencebetweentaxandbookincome—approxi-mately$154.
4billion,or33.
7percentoftaxincome—cannotbeaccountedforbythesehistoricallyrelevantmeasuresofthediscrepancybetweentaxandbookincome.
Toidentifythereasonsforthisbreakdown,thepaperdevelopsamodeloftaxshelteringandthenteststhatmodelbyexploringthedynamicof2Thedifficultiesinisolatingtheseactivitiespreciselyshouldnotbeunderstated.
Aftersur-veyingvariedandconflictingestimatesofthenatureofshelteringactivity,JointCommitteeonTaxation(2000)states,"[T]heJointCommitteestaffbelievesthatdirectmeasurementofcorporatetaxshelteractivitythroughmacroeconomicdataisnotpossible.
Instead,amoreinstructiveapproachmaybetoanalyzespecifictaxsheltertransactionsthathavecometolightandevaluatetheireffectoncorporatereceipts"(p.
16).
Whilenotdisagreeingwiththisapproach,Iattemptinthispaperatleasttouncoverevidenceoftheunderlyingtrendinshelteringactivityandsomeevidenceoftheoverallscopeatthemacroeconomicleveltoinformthedebate.
172Desaithelinkbetweentaxincomeandbookincomeovertimeandbylevelsoftaxincome.
Theestimatesmotivatedbythismodelofshelteringprovideevidencethatthepatternsofthedeterioratinglinkbetweentaxandbookincomeareconsistentwithincreasedlevelsofshelteringoverthedecade.
Oneimportantalternativehypothesis—thatthebreakdownofthelinkbetweentaxincomeandbookincomereflectscoincidentincreasedlevelsofearningsmanagement—isalsoinvestigated.
Measuresassociatedtradi-tionallywithdifferentlevelsof,ormotivesfor,earningsmanagementdonotappeartoexplainfullythedistinctivenatureofthebreakdownofthelinkbetweentaxandbookincomeinthelatterpartofthedecade.
Takentogether,theevidencesuggeststhatthelarge,unexplainedgapsbetweentaxandbookincomethathavearisenduringthelate1990sareatleastpartlyassociatedwithincreasedshelteringactivity.
Section2reviewsalternativemethodologiesforunderstandingthedy-namicsofcorporateprofitsandtheirrelativemeritsandrecentfindings.
Section3examinesthechangingrelationshipbetweentaxincomeandbookincomeasviewedthroughthelensofaggregateddatageneratedfromtaxformswherefirmsreconciletaxandbookincome.
Section4re-viewsindetailthechangingnatureofemployeecompensationanditsimpactonthecorporateprofitbasebyexploitingseveraldifferentsourcesofdataonoptionexercises.
Section5ofthepaperdevelopsamodeloftax-shelteringactivityandthenteststhatmodelthroughanalysesofbookandtaxincomegeneratedfromaccountingstatementsoverthe1990s,withspecialattentionpaidtothealternativeexplanationofearningsman-agement.
Section6presentstheconclusion.
2.
PREVIOUSWORKANDREVIEWOFALTERNATIVEMETHODOLOGIESEffortstounderstandthedynamicsofcorporateprofitsandtaxrevenuestypicallyrelyonthreedistinctsources:(1)economy-wideaggregatedatatakenfromnationalincomeaccounts,(2)reportedtaxandbookincomefromtaxforms,and(3)reportedbookincomeandsimulatedtaxincomefromaccountingstatements.
Eachofthesesourceshasdistinctadvantagesanddisadvantagesinunderstandingthescopeoftax-shelteringactivityand,unsurprisingly,eachyieldsdifferentconclusions.
Thesedistinctsourcesandmethodologiesalsoprovidemuchconfusionabouttheircon-flictingimplications.
Nationalincomeaccountsprovidethemostaggregatepictureofcorpo-rateprofitabilityandallowforadisentanglingofthedifferentreasonswhyaveragetaxratesdepartfromstatutoryrates.
ByimplementingtheTheDivergenceBetweenBookIncomeandTaxIncome173procedureproposedinAuerbachandPoterba(1987),Mackie(2000)pro-videsthisperspectiveonrecentcorporateprofitabilityandtheimplica-tionsforrecenttaxcollections.
Hedocumentsareducedaveragetaxrateamidrisingcorporateprofitabilityandtracesseveralrationalesforthisphenomenon.
Mackieconcludesthatthisaggregateperspectivecannotilluminatethenatureoftax-shelteringactivityasthebasicincomemea-sureemployedpreventsanycorrespondingmeasureofwhatincomewouldhavebeenintheabsenceofshelteringactivity.
Inotherwords,typicalsheltersreducebothincomeandtaxes,leavingaveragetaxratesunaffected.
3Tounderstandthescopeofshelteringactivity,thejointreportingofbookincomeandtaxincomeaffordsthepromiseofmeasuringactivityreportedtoshareholdersbutnotreportedtotaxauthorities.
Indeed,muchoftherecentconcernovertaxsheltersreflectstheuseofTreasurydatatoconstructcomparisonsbetweentaxincomeandbookincomeasre-portedinScheduleM-l.
Thejointreportingofbookincomeandtaxin-comebyfirmsintheirtaxformsaffordsthepossibilityofanalyzingthegapbetweenthetwonotionsofincomeandthevarieddeterminantsofthatgap.
ThereconciliationbetweenbookandtaxincomeinScheduleM-lisoflimiteddetailand,consequently,parsingoutalternativeexpla-nationsofthegapisdifficult.
Finally,theuseofthemicrodataislimitedgivenaccessibilityandconfidentialityrequirementsimposedbytheIRS.
4Nonetheless,theaggregateperspectiveaffordedbythisreporting,alongwiththeabilitytoanalyzetruetaxincomeasopposedtosimulatedtaxincome,makesthisavaluablesource.
Accountingstatementscanalsobeusedtogeneratecomparisonsbe-tweenbookincomeandsimulatedtaxincome.
Inparticular,thereisanextensiveliteraturegaugingthereliabilityofalternativeestimatesoftaxincomegeneratedfromaccountingstatementsbyexplicitlycomparingthemtoIRSdataontaxincome.
5Thesestudiesthentypicallygoontostudythedifferencesbetweenbookandtaxincome,asgeneratedbytheaccountingstatements,tostudythenatureofpermanentandtemporary3Forarelatedeffort,seePetrick(2001)foracomparisononNIPAprofitswithS&P500profits,andadiscussionofthecompositionalreasonswhythesemeasuresmaydiffer.
4Plesko(2002)providesadetailedreconciliationfor1996-1998.
UsingadistinctmeasureofpretaxbookincomeandtaxnetincomefromTalisman(1999)andtheoneemployedinthispaper,hefindsthatthedifferencebetweenpretaxbookincomeandtaxnetincomegrewby71.
9percentbetween1996and1998.
Hismeasureoftheaggregatedifferencebe-tweenpretaxbookandtaxincomeasreportedbyfirmsis$159billionin1998.
5SeePlesko(2000,2003)andMillsandNewberry(2001)foradiscussionofthesealternativemeasuresandtheirrelativemerits.
174Desaidifferencesintaxaccounting.
Theseestimatesofbookandsimulatedtaxincomecanalsobeemployedtoinfersomethingabouttax-shelteringac-tivity,asinManzonandPlesko(2001).
ManzonandPlesko(2001)studythegapbetweenaccounting-baseddefinitionsoftaxincomeandbookincome,anddemonstratethatafewmeasuresapproximatingthedemandfortaxsheltershelpexplainthecross-sectionalvariationinthesegaps.
6AsnotedbyHanlonandShevlin(2001),theestimatesofbookandtaxincomegeneratedbyaccountingstatementsdonotencompassthesamedifferencesasthosegeneratedbyviewingdifferencesbetweenbookandtaxincomeasreportedontaxforms.
Inparticular,deductionsassociatedwiththeexerciseofstockoptionswillnotberepresentedinthegapbe-tweenbookandtaxincomegeneratedfromaccountingstatementsbutwillshowupinthegapgeneratedusingtaxforms.
7Inadditiontothisdistinctioninthereportingofdeductionsassociatedwithoptionexercises,anyanalysisofbookincomemustaddresspotentialmanagerialmotivestomanageearnings.
Manyauthorshavecontributedtotheliteraturethattriestodefine,document,andunderstandthemotivationbehindearningsmanagement.
8Whilethemagnitudeandimpactofsuchbehaviorremainsinquestion,itisconceivablethattrendsinearningsmanagementcouldresultinsystematicvariationinbookandtaxincome.
TheanalysisthatfollowsattemptstousetheIRSandaccountingsourcesoftaxandbookincometoilluminatetheproblemoftaxshelters.
Theaggregateview,asrepresentedbyIRSdata,isanalyzedinsection3intandemwithanexplorationoftheimpactofESOsinsection4.
Thedisaggregatedviewgeneratedbyaccountingstatementsisanalyzedinsection5bydevelopingamodeloftaxshelteringandthentestingit.
Thisanalysisalsodiscussesthepossibilitythatearningsmanagementmightresultinsimilarempiricalpatternsandtriestodistinguishshelteringac-tivityfromearningsmanagement.
6ManzonandPlesko(2001)relatetheabsolutemagnitudeofthegaptopossibledetermi-nantsofdemandforshelteringactivity.
Theyfindthatdummyvariablesassociatedwiththepresenceofpretaxprofitsandnetoperatinglosseshelppredictthesizeofthegapinacross-sectionalregression.
7Thisdifferenceresultsfromthedeparturefrom"cleansurplus"accountingintheaccount-ingofstockoptions.
Inshort,taxbenefitsfromexerciseofESOsdonotflowthroughtheincomestatementbutaresimplytransferredtoadditionalpaid-incapital.
Asaconsequence,taxexpensesfromaccountingstatementswon'tallowforconsiderationofthedeductionsassociatedwithexercises.
Thereportingofthesetaxbenefitsassociatedwithemployeeexer-cisesiscontinuingtoevolve,andHanlonandShevlin(2001)arguethatvariedestimatesoftaxratesmakethemistakeofinferringtaxrateswithoutconsiderationfortheeffectsofthesetaxbenefits.
8SeeDechowandSkinner(2000),HealyandWahlen(1999),andSchipper(1989)forreviewarticlesofthisliterature.
TheDivergenceBetweenBookIncomeandTaxIncome1753.
TAXINCOME,SIMULATEDBOOKINCOME,ANDACTUALBOOKINCOMEConcernsoverincreasedshelteringactivitiesbycorporationsaretypicallyassociatedwithtrendsincorporatetaxreceiptsandtrendsinthegapbetweentaxincomeandbookincomeemployingfiguresextractedfromtaxreturns.
Figure1considerstheratiooffederalcorporatetaxreceiptstoallfederalon-budgettaxreceiptsfrom1971to2001,asreportedinOfficeofManagementandBudget(2002).
Thefiguredepictstheoveralldeclineinthatratio,fromnearly19.
7percentin1977to10.
2percentin2001.
Muchofthepatterninthisratiofortheinterveningyearscanbeexplainedbylargelegislativechangesandgeneraleconomicconditions.
Therecentdecline,beginningin1996,ismorepuzzlinggiventhecoinci-denteconomicexpansion.
Inparticular,thisratiofallsfrom15.
8percentin1996to13.
5percentin2000andthen,evenmoredramatically,to10.
2percentin2001.
TherelationshipbetweentaxandbookincomeasreportedinScheduleM-ldemonstratesasimilarlycurioustrend.
Thesefigureshavebeenem-ployedbytheTreasuryDepartmenttoemphasizetheincreasedincidenceandmagnitudeofcorporatetaxshelteractivity.
Figure2reproducesafigurefromTalisman(1999)thatisalsorelatedtothefigureemployedintheTreasuryDepartment'sstudyofcorporatetaxshelters(U.
S.
Treasury,FIGURE1.
CorporateTaxReceiptsasPercentageofTotalOn-BudgetReceipts,1971-2001251971197619811986199119962001Thegraphshowstheratioofcorporatetaxreceiptstoallon-budgetfederalreceipts,asreportedinOfficeofManagementandBudget(2002).
276DesaiB-BIIPQWDOSOONONXoX.
39xo120,pptooo3tax-ereaitaxeslessthassetsa^oJ§*-ooooOOocgg.
gX3>oSoT3cteen"a,!
CoTheDivergenceBetweenBookIncomeandTaxIncome1771999).
Thisgraphemploysdataforonlythosecorporationswithassetsgreaterthan$1billion.
Inferringamultipleofbookincomeovertaxin-comefromthatgraphillustratestheconcernthattax-shelteringactivityhasincreasedoverthatperiod.
Inparticular,theratioofbookincometotaxincomegrowsto1.
4from1.
0overfiveyears.
Alternatively,bookin-comeexceedstaxincomebyapproximately$120billion(in1992dollars)by1996.
Thisfigure,andthegapithighlights,hasservedasthemostimportantsourceofdataforthedebateoncorporatetaxshelters.
AspointedoutbyKies(1999a),however,thereareseveralalternativeexpla-nationsforthisgapthatmakesuchapictureinconclusive.
Tounderstandtheimplicationofthiswideninggapforthescopeoftax-shelteringactivity,itisusefultoconstructameasureofsimulatedbookincomethatincorporatesidentifiablesourcesofthatgap.
Inparticu-lar,discrepanciesbetweentaxandbookincomemaybeattributabletocausesunrelatedtotax-shelteringactivity.
TheevidenceinTable1andFigure3considersthedynamicsoftaxandbookincomeforfirmsthathaveassetsgreaterthan$250millionandisolatestheimpactstemmingfromthreepotentialsourcesofthatgap:(1)thedifferentialtreatmentofdepreciationexpenseontaxandbookforms,(2)reinvestedearningsabroad,and(3)deductionsassociatedwiththeexerciseofnonqualifiedstockoptions.
Beforeanalyzingthedata,thedistinctionsbetweenthetwosetsofdatashouldbeemphasized.
Figure2employsactualtaxreturndatafrom1991to1996toanalyzethegapbetweenactualtaxandbookincomeforcorporationsthathavemorethan$1billioninassetsandarenotS-corporations,RICs,orREITs.
Incontrast,Table1andFigure3ini-tiallysimulatethatgapforsimilarcorporations,butwithacutoffof$250millioninassetsovertheperiodfrom1982to2000.
Table1providesdata,incurrentdollars,forsimulatingbookincomeforthesecorporationsandbeginswithanestimateoftaxincomecompa-rabletothebaseintheTalismanfigure.
Toanalyzetheimpactofthediscrepancybetweendepreciationexpenseallowedfortaxpurposesandthatassociatedwithbookaccounting,Table1drawsonBureauofEco-nomicAnalysis(BEA)estimatesofthecapitalconsumptionallowance(CCA)adjustment,whichmeasuresthediscrepancybetweentaxmea-suresofdepreciationandeconomicdepreciation.
Thisaggregatemeasureofthediscrepancyisscaledbythatfractionofdepreciationexpensesforfirmswithover$250millioninassets.
Tothedegreethateconomicdepre-ciationisnotrepresentativeofdepreciationassociatedwithhistoriccostaccounting,andtothedegreethatthislinkbetweenaccountingandeco-nomicdepreciationhasbecomemoretenuousovertimegiventhechang-ingnatureofassets,theevidenceinTable1likelyunderstatestheimpactassociatedwithgapsbetweentaxandaccountingnotionsofdepreciation.
178DesaiTABLE1TaxIncomeandComponentsofSimulatedBookIncome,1982-2000*Year1982198319841985198619871988198919901991199219931994199519961997199819992000Taxincome85,223103,103130,569142,224136,194214,618306,665248,551227,723215,572251,587307,302357,936436,533485,645513,332457,575457,575457,575Excessdepreciation8,43719,34030,27643,55733,45731,91530,34123,75913,5427,0313,0142,8668,54513,09517,42723,66027,35637,21831,243Reinvestedearningsabroad4,57413,10716,80313,3979,24717,68713,44711,77520,48817,56815,73329,21223,34245,70846,03547,72831,62058,80684,111ProceedsfromoptionexercisesN/AN/AN/AN/AN/AN/AN/AN/AN/AN/A14,08615,36510,41617,64932,41242,62673,59874,832106,265Simulatedbookincome98,234135,550177,649199,179178,899264,220350,453284,085261,753240,171284,419354,744400,240512,985581,519627,346590,150628,431679,195*Alldollarfiguresareincurrentdollars.
Taxincomeistotalreceiptslesstotaldeductionsforfirmswithgreaterthan$250millioninassets,excludingRICs,REITs,andS-corporations.
Thetaxincomefiguresfor1999and2000arenotavailable,sothesimulatedbookincomefigureassumesthattaxincomestaysat1998levels.
ExcessdepreciationcorrespondstotheCCAadjustmentscaledfortherelativeuseofdepreciationbyfirmswithgreaterthan$250millioninassets.
Reinvestedearningsabroadarethediffer-encebetweenforeignearningsandrepatriationsscaledfortherelativelevelsofforeignincomeforfirmswithgreaterthan$250millioninassets.
Optionexercisesaretheproceedsfromexercises,asmorefullyreportedinTable2.
Simulatedbookincomeisthesumofthosefigures.
Similarly,Table1capturesthediscrepancybetweenincomeearnedbyU.
S.
corporationsoperatingabroadandtheincomethatisrepatriatedfromBEAdataoncapitalflows.
9Finally,Table1providesdatafromExecucompontheaggregatelevelofoptionexercisesbyemployees,whichisdiscussedingreaterdetailbelow.
Thesedataareavailableonlyforyearssubsequentto1992.
Taxincomefrom1999and2000isnotyetavailablebutispresumedtostayat1998levelstofacilitateabaselinecomparisonfor1999and2000.
Figure3relatesthefindingsofTable1inaformatcomparabletotheevidenceprovidedinFigure2bytranslatingthefiguresfromTable19Thesefiguresarescaledbytheannualfractionofforeigntaxcreditsattributabletofirmswithassetsmorethan$250million.
TheDivergenceBetweenBookIncomeandTaxIncome179COsuiCO"^fNiIi^^J^SOv—mov09\ON—soO—OS73>i*55.
S8*S&"tfflS>7-(inElacs2s^^.
s§1SsSi2.
SI-aS>.
2§I-8SIs:is.
gs-agigis**r"pj\—'toWIU,o73":[°^-Stui3^2u->Tfooo"gS23m^§TJ^&HH22-^OIy0cn"OHcn0)G"OHcnCDcncucn0)cn0)"EHcnOJ_C"HHcn(LI_gEHcnEHcnOJUtCUTheDivergenceBetweenBookIncomeandTaxIncome195FIGURE5.
BookIncomeVersusTaxableIncome,SplineEstimatesbyThree-YearIntervalswithIndustryFixedEffects,1992-2000—1992-1994456ln(taxableincome)1995-1997101998-2000ThethreelinesinthefigureareconstructedusingthecoefficientsofsplineregressionsreportedinTable5,whereindustryfixedeffectsareemployedforthegroupsofyears1992-1994,1995-1997,and1998-2000.
FIGURE6.
BookIncomeVersusTaxableIncome,SplineEstimatesbyYearwithIndustryFixedEffects,1992-2000123456ln(taxableincome)199219931994199719981999—o1995-2000101996ThelinesinthefigureareconstructedusingthecoefficientsofsplineregressionsreportedinTable6ofDesai(2002),whereindustryfixedeffectsareemployedforeachyearfrom1992to2000.
196DesaiGiventheindustry-specificityofdiscretionaryaccrualopportunities,itisusefultoemployindustryfixedeffectsinthesepiecewiselinearregres-sionsinanefforttoisolateatax-shelteringexplanationforthisphenome-nonfromanearningsmanagementperspective.
TheresultspresentedinTable5,withtheinclusionofindustryfixedeffects,demonstratethatwithin-industryvariationprovidesevenstrongerevidenceforaweaken-ingoftherelationshipbetweenbookincomeandtaxincomeatlowlevelsoftaxincome.
ThecoefficientsfromthethreespecificationsemployingindustryfixedeffectsareusedtoconstructthelinesinFigure5.
Thisfigureprovidestheflatteningoftherelationshipbetweenbookincomeandtaxincome,aspredictedinthemodelofcostlyshelteringoverthecourseofthe1990s.
Thissameexerciseisrepeatedbyyearwithindustryfixedeffects,andtheresultsarepresentedingraphicalforminFigure6.
Fig-ure6providessimilarresultsbecausethecoefficientsonthefirstsplinearepositiveandsignificantinearlyyearsofthesampleandbecausetheyapproach0by2000.
AsreportedinDesai(2002),theexplanatorypoweroftheregressionsthatunderliesFigure6declinesoverthesampleperiod,furtheremphasizingthereducedlinkbetweenbookincomeandtaxincome.
Thisregressionevidenceisdifficulttoreconcilewithalternativeratio-nalesforthedisparitybetweentaxincomeandbookincome.
Forexample,ifthedifferentialmeasurementofdepreciationorofreinvestedearningsabroadweretoaccountwhollyforthischangedpattern,thisdifferentialmeasurementwouldhavetohavebecomemoreconcentratedamonglow-tax-incomefirmsduringthe1990s.
Whilethesepossibilitiesappearun-likely,itispossiblethattheaccountingofoptionexercises,iffirmsfollowcleansurplusaccounting,wouldhavereducedtaxincome,asmeasuredbyaccountingstatements,foryounggrowthfirmswithlowtaxincome.
Controllingseparatelyforthelevelsofoptionactivityintheseregressionsdemonstratesthatoptionactivitycannotexplainthisdecouplingofbookincomeandtaxincomeatlowlevelsoftaxincome.
Toensurethatresultsarenotdisproportionatelyreflectingthedifferentbehaviorofsmallorlargefirms,appendixtable2andappendixfigureslaandlbofDesai(2002)attemptthesamepiecewiselinearregressionsprovidedinTable5bydividingthesampleatthemedianlevelofsales.
Thesamepatternsappeartohold.
hiadditiontousingindustryfixedeffectstoseparateearningsmanage-mentexplanationsfromtax-shelteringexplanations,itispossibletocon-sidertherelativevariabilityofbookincomeandtaxincomeinanefforttosegregateactiveearningsmanagersfromnonearningsmanagers.
26To26Suchadistinctionresemblesotherstudiesthatattempttoisolateearningssmoothersbytherelativeabsenceofvariabilityinreportedearnings,asinMyersandSkinner(2001),whostudyfirmswithconsistentincreasesinreportedearnings.
TheDivergenceBetweenBookIncomeandTaxIncome197boingsOu-*ON,AON.
"°fc3-^i-JSr-HCOOIboICOCOONOINoooincoNOi-H^oNNbcMcoLNONOoocooooinooNoooo0^TticooNcoinco^cMoOi—iGOrHOOooLn10in*CMCMcoINCMOOOOOOOoooooOOOOOrHOOONOOCMNOON"rtlLNrHinCMi-HlN'HHTtlOOOi-H-^T}01mlineNOlinea,(X)LNlineen0001"OHenONlineen*oi—ilineen"01ff!
zOIc^5gS$eggniSU,^H5b13ggog'uOHCJOHoOHOCUTjouc"o-UingS^f"""ou§.
aacatSH'SfuQJJBCnsoCtnM)vaS°-2-OJo5J.
5g"43TJaT3T3^nsto"5MHiSOJ°goi'Cg_^fill198Desaidoso,ratiosofthestandarddeviationsofbookincometotaxincomearecalculatedforallthefirmsinthesample.
Thisratiohasamedianof0.
91,andfirmswitharatiobelowthemedianareconsideredearnings-smoothingfirmsandfirmswitharatioabovethemedianareconsiderednonsmoothers.
Ifearningmanagementwereaprimarydriverofthebreakdownoftherelationshipbetweenbookincomeandtaxincomeatlowlevelsoftaxincome,thenfirmswithdifferentpropensitiestosmoothbookincomeshouldexhibitadifferentrelationshipbetweentaxincomeandbookincomeatlowlevelsoftaxincomeoverthecourseofthedecade.
TheevidenceprovidedinTable6andFigures7and8suggestotherwisebecausebothearningssmoothersandnon-earningssmoothersexhibitsimilarunderlyingtrendsoverthecourseofthedecadewithrespecttoaflatteningofthebookincome-taxincomerelationshipatlowlevelsoftaxincome.
TheregressionframeworkexploredinTables5and6andFigures5to8providesevidencethatisconsistentwiththemodeloftaxshelteringFIGURE7.
BookIncomeVersusTaxableIncome,Earnings-SmoothingFirms,SplineEstimatesbyThree-YearIntervalswithIndustryFixedEffects,1992-200010-345678910ln(taxableincome)—1992-19941995-1997-X-1998-2000ThethreelinesinthefigureareconstructedusingthecoefficientsofsplineregressionsreportedinTable6forthegroupsofyears1992-1994,1995-1997,and1998-2000.
Thefigureemploysregressionsfortwosamplesoffirmsthataredividedatthemedianratioofthestandarddeviationofbookincometothestandarddeviationoftaxableincome.
Firmswithratiosabovethemedianratioaretermednonearnings-smoothingfirms,andthosewithratiosbelowthemedianratioaretermedearnings-smoothingfirms.
TheDivergenceBetweenBookIncomeandTaxIncome199FIGURE8.
BookIncomeVersusTaxableIncome,Nonearnings-SmoothingFirms,SplineEstimatesbyThree-YearIntervalswithIndustryFixedEffects,1992-20000+-012^—1992-19943456ln(taxableincome)1995-19977->8e—1998-2000ThethreelinesinthefigureareconstructedusingthecoefficientofsplineregressionsreportedinTable6forthegroupsofyears1992-1994,1995-1997,and1998-2000.
Thefigureemploysregressionsfortwosamplesoffirmsthataredividedatthemedianratioofthestandarddeviationofbookincometothestandarddeviationoftaxableincome.
Firmswithratiosabovethemedianratioaretermednonearnings-smoothingfirms,andthosewithratiosbelowthemedianratioaretermedearnings-smoothingfirms.
providedinsection5.
2.
Theimplicationoftheseresultsisthatshelteringbecameconsiderablylesscostly,eitherthroughloweredprobabilitiesofdetectionorperceivedlowerpenalties,andthatfirmsbecamemoreag-gressiveduringthe1990s.
Thealternativeexplanationofasecularin-creaseinearningsmanagementisdifficulttoreconcilewiththedataprovidedinthepaper.
First,earningsmanagementtheoriestypicallydonotallowforsuchlong-runintertemporalshiftingofincome,aswouldberequiredtoexplaintheaggregatetrendsinFigure3.
Second,ifearningsmanagementopportunitiesareparticularlyassociatedwithdiscretionaryaccrualopportunitiesinsomeindustries,thenwithin-industryvariationshouldhavereduced,ratherthanincreased,evidenceofthisunderlyingbehaviorinthemicroevidence.
Finally,ifincreasedearningsmanage-mentwerethecauseoftheflatteningoftherelationshipbetweenbookincomeandtaxincomeatlowlevelsoftaxincome,simpledistinctionsintherelativevariabilityofbookincomeandtaxincomeshouldhavedemonstratedmorepronounceddifferencesinthisbreakdown.
Whileitisnotpossibletoruleoutearningsmanagementdefinitively—particularly200Desaifraudulentbookreportingofincome—asasourceofsomeoftheaggre-gatephenomenaobservedinFigure3,thismicroanalysissuggeststhatthedistinctivewayinwhichtherelationshipbetweenbookincomeandtaxincomehasdeterioratedoverthedecadeisconsistentwithincreasedlevelsofcostlysheltering.
6.
CONCLUSIONThispaperattemptstoilluminatethedebateontaxshelteringbydisentan-glingvariedexplanationsforthebreakdownintherelationshipbetweentaxincomeandbookincomeoverthelastdecade.
First,thepaperdemon-stratesthatidentifiablefactorsthatonceaccountedforthedifferencebe-tweentaxincomeandbookincomefailtoaccountforthelargegapbetweenbookincomeandtaxincomeinmorerecentyears.
Inparticular,by1998,morethanhalfofthedifferencebetweentaxincomeandbookincome—approximately$154.
4billion,or33.
7percentoftaxincome—cannotbeaccountedforbythesehistoricallyrelevantmeasuresofthediscrepancybetweentaxincomeandbookincome.
Second,thepaperdemonstratesthatwithinthoseidentifiablefactors,changedpatternsinemployeecompensationarecreatingthelargestdisparitybetweenbookincomeandtaxincome.
By1998,differenttreatmentofemployeeoptionsaccountedfornearlythreetimesasmuchofthedifferencebetweenbookincomeandtaxincomeasdiddistinctdefinitionsofdepreciation.
Byexaminingthissamedisparitybetweenbookincomeandtaxincomeusingaccountingdata,thispaperdemonstratesthatthisbreakdownhasbeenparticularlypronouncedatlowlevelsoftaxincomeandhasdegen-eratedprogressivelyoverthedecade.
Theseresultsareconsistentwiththemodelofcostlytaxshelteringpresentedinthepaperanddecreasedcostsofshelteringoverthedecadeleadingtogreatergapsbetweentaxincomeandbookincome.
Teststocheckiftheseresultsaredrivenbyincreasedlevelsofearningssmoothingdonotappeartodiminishthere-sultsthatsuggestthatincreasedtaxshelteringisresponsibleforthedis-tinctivebreakdownintherelationshipbetweenbookincomeandtaxincome.
ThelargediscrepancyreportedbetweensimulatedbookincomeandactualbookincomefromFigure3,alongwiththeregressionevidence,suggeststhateffortsbyfirmstocircumventtaxpaymentsarebecomingmoresignificant,cheapertoimplement,andhardertodetect.
Thesedevel-opmentsprovideyetanotherreasontoreevaluatethemannerinwhichcorporateearningsaretaxedbecausetheunderlyingdevelopmentsdriv-ingthesephenomena—includingincreasedaccesstoglobalopportunitiesandtherapiddevelopmentoffinancialinnovations—areunlikelytode-TheDivergenceBetweenBookIncomeandTaxIncome201clineinimportanceinthenearfuture.
Similarly,thesetrendsillustratetheneedtorevisittherationalesfordistinctivebookandtaxreporting,andthedegreeandmannerinwhichfirmsareforcedtoreconcilebookincomeandtaxincome.
APPENDIXInordertomotivatetheempiricaltestsofincreasedtaxsheltering,itisusefultospecifyamodelofcostlytaxsheltering.
LetIequaltrueeconomicincome,Bequalreportedbookincome,andTequalreportedtaxincome.
Bookincomeispresumedtocorrespondtotrueeconomicincomewithnoise,sothat:B=Ieuwhereu~N(0,a2)(1)Tospecifywhattaxincomewouldbe,letsequalshelteredincome,whichisassociatedwithcostsy(s),whichinturnarecharacterizedbyy'(s)>0andy"(s)>0.
Inthissetting,thecostsofshelteringcorrespondonlytotheamountofincomeshelteredandnottheamountoftrueeco-nomicincome.
Inotherwords,largerfirmshavenoadvantageinshelter-ingincomerelativetosmallerfirms.
Asaconsequence,taxincomeisgivenby:T=I-s-y(s)(2)andthetaxrateisafunctionofthistaxableincome,asrepresentedbyx[I-s-y(s)].
Firmschoosetheamountofincometoshelterbysolving:max[I-s-y(s)]{l-T[I-s-y(s)]}+s(3)swhichyieldsthefirst-ordercondition:-[1+y'(s)][l-T(-)]+[I-s-y(s)][l+y'(s)]T'O+1=0(4)Toanalyzethisproblemfurther,itisusefultoconsidertworegionsofthetaxschedule:onewherethereisprogressivityandonewherethere202Desaiisnoprogressivity.
Atsufficientlyhighlevelsofincome,thetaxratewillexhibitnoprogressivity;T'Q=0,andequation(4)willcollapseto:[1+y'(s)][l-T0]=1(5)Atsuchlevelsofincome,levelsofshelteringwillnolongervarywithlevelsofincomebecause:dlAsaconsequence,itispossibletorewritetheleveloftaxincomeinequa-tion(2)as:T=I-s-y(s)=I-k(6)Takinglogarithmsofbothsidesofequations(6)and(1)andemployingafirst-orderTaylorapproximation,itispossibletorewriteequation(1)as:ln(B)=ln(I)+-+u(7)whichinturnbecomestheestimatingequationintheempiricalanalysisthatfollows.
Whenlevelsofincomearesuchthatx'(-)^0,thefirst-orderconditioninequation(4)canberewrittenasanexpressionfortaxincome:(1-T)-Y(S)](8)Tomakethisequationtractable,itisusefultotransformitinto:1+y'(s)(9)whereF(x)=[T'(-)X+x(-)—1],andtodefinetheinversefunctionasG(z)=F~l(z).
Thisspecificationallowstaxincometoberewrittenas:TheDivergenceBetweenBookIncomeandTaxIncome203T=G—(10)\l+Y(s)JThecurvatureofG(-)willbesmallwhenthecurvatureofF(-)ishigh,soitbecomesclearthatatlowlevelsofincome=0.
dlWithrespecttotheestimatingequationprovidedinequation(7),thismodelsuggeststhattherelationshipbetweenbookincomeandtaxin-comewillbedistinctiveatlowlevelsoftaxincome,necessitatingapiecewiselinearmodeltoestimateequation(7).
Morespecifically,themodelsuggeststhattaxshelteringwillbereflectedbyarelativelyflatterrelationshipbetweenbookincomeandtaxincomeatlowlevelsoftaxincome.
Correspondingly,increasedlevelsoftaxshelteringwillbere-flectedinaflatteningoftherelationshipbetweentaxincomeandbookincome.
REFERENCESAdams,J.
(2001).
"APensionAccountingPrimer.
"CSFBEquityResearchReport.
June.
Auerbach,A.
,andJ.
Poterba(1987).
"WhyHaveCorporateTaxRevenuesDe-clined"InTaxPolicyandtheEconomy,vol.
1,L.
Summers(ed.
).
Cambridge,MA:MITPress.
Bankman,J.
(1999).
"TheNewMarketinCorporateTaxShelters.
"TaxNotesToday83:1775-1795.
Barnea,A.
,J.
Ronen,andS.
Sadan(1975).
"TheImplementationofAccountingObjectives—AnApplicationtoExtraordinaryItems.
"TheAccountingReview50:58-68.
BearStearns(2000).
"EmployeeStockOptions.
"AccountingIssues(July6).
Brenner,M.
,R.
Sundaram,andD.
Yermack(2000).
"AlteringtheTermsofExecu-tiveStockOptions.
"JournalofTinancialEconomics57:103-128.
Burgstahler,D.
,andI.
Dichev(1997).
"EarningsManagementtoAvoidEarningsDecreasesandLosses.
"JournalofAccountingandEconomics24(1):99-126.
Callihan,D.
(1994).
"CorporateEffectiveTaxRates:ASynthesisoftheLiterature.
"JournalofAccountingLiterature13:1-43.
27Moreprecisely,thisequationrequirestheadditionalassumptionthat%"{)=0andthaty"(-)^T'Obecausef(s)f1dly"(s)+2T'Q[1+Y'(S)]3204DesaiCore,J.
,andW.
Guay(2001).
"StockOptionPlansforNon-ExecutiveEmployees.
"JournalofFinancialEconomics61(2):253-287.
Dechow,P.
,andD.
Skinner(2000).
"EarningsManagement:ReconcilingtheViewsofAccountingAcademics,Practitioners,andRegulators.
"AccountingHorizons14(2):235-250.
Degeorge,F.
,J.
Patel,andR.
Zeckhauser(1998).
"EarningsManagementtoExceedThresholds.
"JournalofBusiness72(l):l-33.
Desai,M.
(2002).
"TheCorporateProfitBase,TaxShelteringActivity,andtheChangingNatureofEmployeeCompensation.
"NBERWorkingPaperno.
8866.
March.
Desai,M.
,C.
Foley,andJ.
HinesJr.
(2001).
"RepatriationTaxesandDividendDistortions.
"NationalTaxJournal54:829-851.
Desai,M.
,C.
Foley,andJ.
HinesJr.
(2003).
"ChainsofOwnership,RegionalTaxCompetition,andForeignDirectInvestment.
"InForeignDirectInvestmentintheRealandFinancialSectorofIndustrialCountries,HeinzHerrmannandRobertLip-sey(eds.
).
Heidelberg,Germany:Springer-Verlag.
Desai,M.
,andJ.
HinesJr.
(2002).
"ExpectationsandExpatriations:TracingtheCausesandConsequencesofCorporateInversions.
"NationalTaxJournal55(3):409-440.
Dworin,L.
(1985).
"OnEstimatingCorporateTaxLiabilitiesfromFinancialState-ments.
"TaxNotesToday85:965-971.
Dye,R.
(1988).
"EarningsManagementinanOverlappingGenerationsModel.
"JournalofAccountingResearch26(2):195-235.
Engel,E.
,M.
Erickson,andE.
Maydew(1999).
"Debt-EquityHybridSecurities.
"JournalofAccountingResearch37(2):249-274.
Fudenberg,D.
,andJ.
Tirole(1995).
"ATheoryofIncomeandDividendSmoothingBasedonIncumbencyRents.
"JournalofPoliticalEconomy103(l):75-93.
Goel,A.
,andA.
Thakor(2003).
"WhyDoFirmsSmoothEarnings"ForthcominginJournalofBusiness76(1).
Goolsbee,A.
(2000a).
"WhatHappensWhenYouTaxtheRichEvidencefromExecutiveCompensation.
"JournalofPoliticalEconomy108(2):352-378.
(2000b).
"Taxes,High-IncomeExecutives,andthePerilsofRevenueEsti-mationintheNewEconomy.
"TheAmericanEconomicReview90(2):271-275.
Hall,B.
,andJ.
Liebman(2000).
"TheTaxationofExecutiveCompensation.
"InTaxPolicyandtheEconomy,vol.
14,J.
Poterba(ed.
).
Cambridge,MA:MITPress.
Hanlon,M.
,andT.
Shevlin(March2002).
"AccountingforTaxBenefitsofEm-ployeeStockOptionsandImplicationsforResearch.
"PublishedinAccountingHorizons16:1-16.
Healy,P.
,andJ.
Wahlen(1999).
"AReviewoftheEarningsManagementLitera-tureandItsImplicationsforStandardSetting.
"AccountingHorizons13(4):365-383.
Huddart,S.
,andM.
Lang(1996).
"EmployeeStockOptionExercises:AnEmpiricalAnalysis.
"JournalofAccountingandEconomics21:5-43.
JointCommitteeonTaxation(2000).
"TestimonyoftheStaffoftheJointCommit-teeonTaxationConcerningInterestandPenaltiesandCorporateTaxSheltersBeforetheSenateCommitteeonFinance.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
March.
Jones,J.
(1991).
"EarningsManagementDuringImportReliefInvestigations.
"JournalofAccountingResearch29:193-228.
Kies,K.
(1999a).
"StatementofKennethKies:TestimonyBeforetheHouseCom-TheDivergenceBetweenBookIncomeandTaxIncome205mitteeonWaysandMeansHearingonPenaltyProvisioninPresident'sFiscalYear2000Budget.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
March.
(1999b).
"ACriticalLookattheAdministration's'CorporateTaxShelter'Proposals.
"TaxNotesToday83:1463-1486.
(1999c).
"StatementofKennethKies:TestimonyBeforetheHouseCommit-teeonWaysandMeansHearingonCorporateTaxShelters.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
November.
(2000).
"StatementofKennethKies:TestimonyBeforetheSenateFinanceCommitteeHearingonPenaltyandInterestProvisionsintheInternalRevenueCode.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
March.
Kinney,M.
,andE.
Swanson(1993).
"TheAccuracyandAdequacyofTaxDatainCompustat.
"JournalofAmericanTaxationAssociation15:121-135.
Mackie,].
,Ill(2000).
"TheComebackoftheCorporateIncomeTax.
"InProceedingsoftheNinety-SecondAnnualConferenceonTaxation.
Washington,D.
C.
:NationalTaxAssociation,93-102.
Manzon,G.
,Jr.
,andG.
Plesko(2001).
"TheRelationBetweenFinancialandTaxReportingMeasuresofIncome.
"TaxLawReview55(2):175-214.
Mclntyre,R.
(2000).
"CorporateIncomeTaxesinthe1990s.
"Washington,D.
C.
:InstituteonTaxationandEconomicPolicy.
October.
Miller,G.
,andD.
Skinner(1998).
"DeterminantsoftheValuationAllowancesforDeferredTaxAssetsUnderSFASNo.
109.
"TheAccountingReview73(2):213-233.
Mills,L.
,andK.
Newberry(2001).
"TheInfluenceofTaxandNon-TaxCostsonBook-TaxReportingDifferences:PublicandPrivateFirms.
"TheJournaloftheAmericanTaxationAssociation23(1).
Murphy,K.
(1999).
"ExecutiveCompensation.
"InHandbookofLaborEconomics,vol.
Ill,OrleyAshenfelterandDavidCard(eds.
).
NewYork:North-Holland.
Myers,L.
,andD.
Skinner(2001).
"EarningsMomentumandEarningsManage-ment.
"UniversityofMichiganWorkingPaper.
OfficeofManagementandBudget(2002).
"HistoricalTables:FiscalYear2003BudgetoftheU.
S.
Government.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
Omer,T.
,K.
Molloy,andD.
Ziebart(1991).
"MeasurementofEffectiveCorporateTaxRatesUsingFinancialStatementInformation.
"JournaloftheAmericanTaxa-tionAssociation13:57-72.
Petrick,K.
(2001).
"ComparingNIPAProfitswithS&P500Profits.
"SurveyofCur-rentBusiness(April):16-20.
Plesko,G.
(2000).
"Book-TaxDifferencesandtheMeasurementofCorporateIn-come.
"InProceedingsoftheNinety-SecondAnnualConferenceonTaxation.
Wash-ington,D.
C.
:NationalTaxAssociation.
(2002).
"ReconcilingCorporationBookandTaxNetIncome,TaxYears1996-1998.
"SOIBulletin.
Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
(2003).
"AnEvaluationofAlternativeMeasuresofCorporateTaxRates.
"ForthcominginJournalofAccountingandEconomics,February2003.
Porcano,T.
(1986).
"CorporateTaxRates:Progressive,Proportional,orRegres-sive.
"JournaloftheAmericanTaxationAssociation7:17-31.
Schipper,K.
(1989).
"Commentary:EarningsManagement.
"AccountingHorizons3(4):91-102.
Shevlin,T.
(1987).
"TaxesandOff-BalanceSheetFinancing:ResearchandDevel-opmentLimitedPartnerships.
"TheAccountingReview62(3):480-509.
206DesaiStickney,C,andV.
McGee(1982).
"EffectiveCorporateTaxRates—TheEffectofSize,CapitalIntensity,Leverage,andOtherFactors.
"JournalofAccountingandPublicPolicy1:125-152.
Sullivan,M.
(1999a).
"ShelterFalloutCorporateTaxesDown,ProfitsUp.
"TaxNotesToday84:653-657.
(1999b).
"ARevenueEstimateforCorporateTaxShelters.
"TaxNotesToday85:981-983.
(2000a).
"Lobbyist'sFiguresFlawed,DataIndicateCorporateShortfalls.
"TaxNotesToday86:309-313.
(2000b).
"LettheGoodTimesRoll:OptionsandTax-FreeProfits.
"TaxNotesToday87:1185-1191.
Talisman,J.
(1999).
"StatementofJonathanTalisman:TestimonyBeforetheHouseCommitteeonWaysandMeansHearingonCorporateTaxShelters.
"Washing-ton,D.
C.
:U.
S.
GovernmentPrintingOffice.
November.
Trueman,D.
,andS.
Titman(1988).
"AnExplanationforAccountingIncomeSmoothing.
"JournalofAccountingResearch26:127-139.
U.
S.
DepartmentoftheTreasury(1999).
"TheProblemofCorporateTaxShelters:Discussion,AnalysisandLegislativeProposals.
"Washington,D.
C.
:U.
S.
Gov-ernmentPrintingOffice.
Zimmerman,J.
(1983).
"TaxesandFirmSize.
"JournalofAccountingandEconomics5:119-149.
EditedbyJamesM.
PoterbaNationalBureauofEconomicResearch0-262-66138-OfrelatedinterestTaxPolicyandtheEconomy,Volume16EditedbyJamesM.
PoterbaArticlesbyRobertMoffitt;DavidFiglio;JonathanGruber;BrigitteMadrianandDavidLaibson;RonaldLee;CharlesMcLureTaxPolicyandtheEconomy,Volume15EditedbyJamesM.
PoterbaArticlesbyMihirDesaiandJamesHines,Jr.
;RobertCarroll,DouglasHoltz-Eakin,MarkRider,andHarveyRosen;AndrewMitrusiandJamesM.
Poterba;DavidEllwoodandJeffreyLiebman;MartinFeldsteinandElenaRanguelova;MarkMcClellanTheMITPressMassachusettsInstituteofTechnologyCambridge,Massachusetts02142http://mitpress.
mit.
edu
Poterba,editorVolumePublisher:MITPressVolumeISBN:0-262-16220-2VolumeURL:http://www.
nber.
org/books/pote03-1ConferenceDate:October8,2002PublicationDate:January2003Title:TheDivergencebetweenBookIncomeandTaxIncomeAuthor:MihirA.
DesaiURL:http://www.
nber.
org/chapters/c11538THEDIVERGENCEBETWEENBOOKINCOMEANDTAXINCOMEMihirA.
DesaiHarvardBusinessSchoolandNBEREXECUTIVESUMMARYThispaperexaminestheevolutionofthecorporateprofitbaseandtherelationshipbetweenbookincomeandtaxincomeforU.
S.
corporationsoverthelasttwodecades.
Thepaperdemonstratesthatthisrelationshiphasbrokendownoverthe1990s,andithasbrokendowninamannerconsistentwithincreasedtax-shelteringactivity.
Thepapertracesthegrowingdiscrepancybetweenbookincomeandtaxincomeassociatedwithdifferentialtreatmentsofdepreciation,thereportingofforeignsourceincome,andinparticularthechangingnatureofemployeecom-pensation.
Forthelargestpubliccompanies,proceedsfromoptionexer-cisesequaled27percentofoperatingcashflowfrom1996to2000.
Thesedeductionsappeartobefullyutilized,therebycreatingthelargestdistinc-tionbetweenbookincomeandtaxincome.
Whilethedifferentialtreat-mentoftheseitemshashistoricallyaccountedfullyforthediscrepancybetweenbookincomeandtaxincome,thispaperdemonstratesthatbookandtaxincomehavedivergedmarkedlyforreasonsnotassociatedwiththeseitemsduringthelate1990s.
In1998,morethanhalfthedifferenceIamgratefultoAlanAuerbach,JamesHines,PeterMerrill,LilMills,GeorgePlesko,JamesPoterba,andparticipantsattheInternationalSeminarinPublicEconomicsConferenceandtheTaxPolicyandtheEconomyConferenceforhelpfulcomments.
ResearchassistancefromMarkVeblen,JamesZeitler,andYumingZou,andfinancialsupportfromtheDivisionofResearchatHarvardBusinessSchoolaregratefullyacknowledged.
270Desaibetweentaxandbookincome—approximately$154.
4billion,or33.
7per-centoftaxincome—cannotbeaccountedforbythesefactors.
Thispaperproceedstodevelopandtestamodelofcostlytaxshelteringanddemon-stratesthatthebreakdownintherelationshipbetweentaxincomeandbookincomeisconsistentwithincreasinglevelsofshelteringduringthelate1990s.
Thesetestsalsoexploreanalternativeexplanationofthesere-sults—coincidentincreasedlevelsofearningsmanagement—andfindthatthenatureofthebreakdownbetweenbookandtaxincomecannotbeexplainedfullybythisalternativeexplanation.
1.
INTRODUCTIONRecenttrendsinthecorporateprofitbase,andthetaxrevenuesitgener-ates,havedrawnconsiderableattention.
Inparticular,thegapbetweenbookincomeandtaxincomehasreportedlywidenedoverthelastdecade,andcorporatetaxreceiptshavebeenlowerthanexpected.
Seenagainstthebackdropofaprotractedeconomicexpansionoverthe1990s,thesetrendshavegeneratedconsiderableconcernovertheproliferationoftax-shelteringactivitybycorporations.
Theseconcernshavebeenamplifiedfurtherbyanecdotalevidenceontheproliferationoftheseschemesandbyhigh-profilecasesassociatedwithlargecorporations.
Estimatesoftax-shelteringactivityhaverangedwidely,dependingonthesourceandmethodologyemployedingeneratingthoseestimates.
1Thefocusoncorporatetaxsheltersraisesseveralnewquestionsrelatedtotheintegrityofthecorporatetaxbase.
Inparticular,financialinnova-tionsthatreducethecostsof,andwidenthescopefor,recharacterizingincomemayleadtodifficultiesinmaintainingthecorporatetaxbase.
Sim-ilarly,theincreasedimportanceoflegitimateforeignoperationsforU.
S.
firms,theattendanttransfer-pricingopportunitiesaffordedbysuchoper-ations,andincreasedcompetitionbyjurisdictionsforthoseinvestmentsmayincreasethescopeforincomeshifting.
Inshort,U.
S.
firmsarefacedwithenhancedopportunitiesforavoidingorevadingcorporatetaxesthroughcheaper,moresophisticated,andlesstransparentmechanisms.
Thesequestionsandconcernsovertheviabilityofthecorporatetaxbaseandthescopeofshelteringactivityhavegivenrisetovariedlegisla-tiveproposalsbutlimitedempiricalanalysisoftheclaimsassociatedwithincreasedtax-shelteringactivity.
Additionally,evidenceonshelteringac-1Theconcernsregardingthescopeandincreasedincidenceoftaxshelterscanbefoundinvarioussources,includingBankman(1999),Sullivan(1999a,1999b,2000a)andU.
S.
Treasury(1999).
Talisman(1999)andKies(1999a,1999b,1999c,and2000)providealternativeinterpre-tationsofthesetrends.
Foraperspectiveonthevarietyofinstrumentsassociatedwithshel-teringactivity,seeU.
S.
Treasury(1999).
TheDivergenceBetweenBookIncomeandTaxIncome171tivitythatcomesfromreportedbookincomeandtaxincomeraisestheissueofotherdifferencesthatmightarisebetweenbookandtaxincomeandtheirgrowingincidenceormagnitude.
Thus,empiricalexplorationsofthenatureofshelteringactivitymustjointlyinvestigatethechangingnatureofbookandtaxincometoidentifyanyunderlyingtrendsinthegapsthatseparatethem.
2Thispaperattemptstoilluminaterecenttrendsinthecorporateprofitbasetohighlightthepossibleaggregatescopeofshelteringactivity,theotherdeterminantsofthegapsbetweenbookandtaxincomeandtheirscopeforexplainingrecenttrends,andthedynamicsoftherelationshipbetweentaxandbookincomeoverthelastdecade.
Theevidenceprovidedinthispaperdemonstratesthatthelinkbetweenbookandtaxincomehasbrokendownoverthelastdecadefortworeasons.
First,theidentifi-ablefactorsthathavebeenassociatedtraditionallywiththedistinctionbetweenbookandtaxincomehavegrowntremendously.
Inadditiontothegrowingrelevanceofoverseasoperationsandthedifferentmeasuresofdepreciation,thepaperexplorestheproliferationofemployeestockoptions(ESOs)andquantifiestheirimpactonthecorporateprofitbase.
From1996to2000,netproceedsfromtheexerciseofESOsforthelargestU.
S.
corporationsapproximated27percentofoperatingcashflows.
Thesenetproceedsappeartobeutilizednearlyfullyasdeductionsagainstpre-taxincomeforthecorporationsstudied.
Second,thebreakdownbetweentaxincomeandbookincomeisnotlimitedtothegrowthofthesetraditionaldistinctionsbetweentaxandbookincome.
Whilethedistinctivetreatmentoftheseitemshashistori-callyaccountedfullyforthedifferencebetweenbookandtaxincome,thepaperdemonstratesthatbookandtaxincomehavedivergedmarkedlyforreasonsnotassociatedwiththeseitemsduringthelate1990s.
In1998,morethanhalfofthedifferencebetweentaxandbookincome—approxi-mately$154.
4billion,or33.
7percentoftaxincome—cannotbeaccountedforbythesehistoricallyrelevantmeasuresofthediscrepancybetweentaxandbookincome.
Toidentifythereasonsforthisbreakdown,thepaperdevelopsamodeloftaxshelteringandthenteststhatmodelbyexploringthedynamicof2Thedifficultiesinisolatingtheseactivitiespreciselyshouldnotbeunderstated.
Aftersur-veyingvariedandconflictingestimatesofthenatureofshelteringactivity,JointCommitteeonTaxation(2000)states,"[T]heJointCommitteestaffbelievesthatdirectmeasurementofcorporatetaxshelteractivitythroughmacroeconomicdataisnotpossible.
Instead,amoreinstructiveapproachmaybetoanalyzespecifictaxsheltertransactionsthathavecometolightandevaluatetheireffectoncorporatereceipts"(p.
16).
Whilenotdisagreeingwiththisapproach,Iattemptinthispaperatleasttouncoverevidenceoftheunderlyingtrendinshelteringactivityandsomeevidenceoftheoverallscopeatthemacroeconomicleveltoinformthedebate.
172Desaithelinkbetweentaxincomeandbookincomeovertimeandbylevelsoftaxincome.
Theestimatesmotivatedbythismodelofshelteringprovideevidencethatthepatternsofthedeterioratinglinkbetweentaxandbookincomeareconsistentwithincreasedlevelsofshelteringoverthedecade.
Oneimportantalternativehypothesis—thatthebreakdownofthelinkbetweentaxincomeandbookincomereflectscoincidentincreasedlevelsofearningsmanagement—isalsoinvestigated.
Measuresassociatedtradi-tionallywithdifferentlevelsof,ormotivesfor,earningsmanagementdonotappeartoexplainfullythedistinctivenatureofthebreakdownofthelinkbetweentaxandbookincomeinthelatterpartofthedecade.
Takentogether,theevidencesuggeststhatthelarge,unexplainedgapsbetweentaxandbookincomethathavearisenduringthelate1990sareatleastpartlyassociatedwithincreasedshelteringactivity.
Section2reviewsalternativemethodologiesforunderstandingthedy-namicsofcorporateprofitsandtheirrelativemeritsandrecentfindings.
Section3examinesthechangingrelationshipbetweentaxincomeandbookincomeasviewedthroughthelensofaggregateddatageneratedfromtaxformswherefirmsreconciletaxandbookincome.
Section4re-viewsindetailthechangingnatureofemployeecompensationanditsimpactonthecorporateprofitbasebyexploitingseveraldifferentsourcesofdataonoptionexercises.
Section5ofthepaperdevelopsamodeloftax-shelteringactivityandthenteststhatmodelthroughanalysesofbookandtaxincomegeneratedfromaccountingstatementsoverthe1990s,withspecialattentionpaidtothealternativeexplanationofearningsman-agement.
Section6presentstheconclusion.
2.
PREVIOUSWORKANDREVIEWOFALTERNATIVEMETHODOLOGIESEffortstounderstandthedynamicsofcorporateprofitsandtaxrevenuestypicallyrelyonthreedistinctsources:(1)economy-wideaggregatedatatakenfromnationalincomeaccounts,(2)reportedtaxandbookincomefromtaxforms,and(3)reportedbookincomeandsimulatedtaxincomefromaccountingstatements.
Eachofthesesourceshasdistinctadvantagesanddisadvantagesinunderstandingthescopeoftax-shelteringactivityand,unsurprisingly,eachyieldsdifferentconclusions.
Thesedistinctsourcesandmethodologiesalsoprovidemuchconfusionabouttheircon-flictingimplications.
Nationalincomeaccountsprovidethemostaggregatepictureofcorpo-rateprofitabilityandallowforadisentanglingofthedifferentreasonswhyaveragetaxratesdepartfromstatutoryrates.
ByimplementingtheTheDivergenceBetweenBookIncomeandTaxIncome173procedureproposedinAuerbachandPoterba(1987),Mackie(2000)pro-videsthisperspectiveonrecentcorporateprofitabilityandtheimplica-tionsforrecenttaxcollections.
Hedocumentsareducedaveragetaxrateamidrisingcorporateprofitabilityandtracesseveralrationalesforthisphenomenon.
Mackieconcludesthatthisaggregateperspectivecannotilluminatethenatureoftax-shelteringactivityasthebasicincomemea-sureemployedpreventsanycorrespondingmeasureofwhatincomewouldhavebeenintheabsenceofshelteringactivity.
Inotherwords,typicalsheltersreducebothincomeandtaxes,leavingaveragetaxratesunaffected.
3Tounderstandthescopeofshelteringactivity,thejointreportingofbookincomeandtaxincomeaffordsthepromiseofmeasuringactivityreportedtoshareholdersbutnotreportedtotaxauthorities.
Indeed,muchoftherecentconcernovertaxsheltersreflectstheuseofTreasurydatatoconstructcomparisonsbetweentaxincomeandbookincomeasre-portedinScheduleM-l.
Thejointreportingofbookincomeandtaxin-comebyfirmsintheirtaxformsaffordsthepossibilityofanalyzingthegapbetweenthetwonotionsofincomeandthevarieddeterminantsofthatgap.
ThereconciliationbetweenbookandtaxincomeinScheduleM-lisoflimiteddetailand,consequently,parsingoutalternativeexpla-nationsofthegapisdifficult.
Finally,theuseofthemicrodataislimitedgivenaccessibilityandconfidentialityrequirementsimposedbytheIRS.
4Nonetheless,theaggregateperspectiveaffordedbythisreporting,alongwiththeabilitytoanalyzetruetaxincomeasopposedtosimulatedtaxincome,makesthisavaluablesource.
Accountingstatementscanalsobeusedtogeneratecomparisonsbe-tweenbookincomeandsimulatedtaxincome.
Inparticular,thereisanextensiveliteraturegaugingthereliabilityofalternativeestimatesoftaxincomegeneratedfromaccountingstatementsbyexplicitlycomparingthemtoIRSdataontaxincome.
5Thesestudiesthentypicallygoontostudythedifferencesbetweenbookandtaxincome,asgeneratedbytheaccountingstatements,tostudythenatureofpermanentandtemporary3Forarelatedeffort,seePetrick(2001)foracomparisononNIPAprofitswithS&P500profits,andadiscussionofthecompositionalreasonswhythesemeasuresmaydiffer.
4Plesko(2002)providesadetailedreconciliationfor1996-1998.
UsingadistinctmeasureofpretaxbookincomeandtaxnetincomefromTalisman(1999)andtheoneemployedinthispaper,hefindsthatthedifferencebetweenpretaxbookincomeandtaxnetincomegrewby71.
9percentbetween1996and1998.
Hismeasureoftheaggregatedifferencebe-tweenpretaxbookandtaxincomeasreportedbyfirmsis$159billionin1998.
5SeePlesko(2000,2003)andMillsandNewberry(2001)foradiscussionofthesealternativemeasuresandtheirrelativemerits.
174Desaidifferencesintaxaccounting.
Theseestimatesofbookandsimulatedtaxincomecanalsobeemployedtoinfersomethingabouttax-shelteringac-tivity,asinManzonandPlesko(2001).
ManzonandPlesko(2001)studythegapbetweenaccounting-baseddefinitionsoftaxincomeandbookincome,anddemonstratethatafewmeasuresapproximatingthedemandfortaxsheltershelpexplainthecross-sectionalvariationinthesegaps.
6AsnotedbyHanlonandShevlin(2001),theestimatesofbookandtaxincomegeneratedbyaccountingstatementsdonotencompassthesamedifferencesasthosegeneratedbyviewingdifferencesbetweenbookandtaxincomeasreportedontaxforms.
Inparticular,deductionsassociatedwiththeexerciseofstockoptionswillnotberepresentedinthegapbe-tweenbookandtaxincomegeneratedfromaccountingstatementsbutwillshowupinthegapgeneratedusingtaxforms.
7Inadditiontothisdistinctioninthereportingofdeductionsassociatedwithoptionexercises,anyanalysisofbookincomemustaddresspotentialmanagerialmotivestomanageearnings.
Manyauthorshavecontributedtotheliteraturethattriestodefine,document,andunderstandthemotivationbehindearningsmanagement.
8Whilethemagnitudeandimpactofsuchbehaviorremainsinquestion,itisconceivablethattrendsinearningsmanagementcouldresultinsystematicvariationinbookandtaxincome.
TheanalysisthatfollowsattemptstousetheIRSandaccountingsourcesoftaxandbookincometoilluminatetheproblemoftaxshelters.
Theaggregateview,asrepresentedbyIRSdata,isanalyzedinsection3intandemwithanexplorationoftheimpactofESOsinsection4.
Thedisaggregatedviewgeneratedbyaccountingstatementsisanalyzedinsection5bydevelopingamodeloftaxshelteringandthentestingit.
Thisanalysisalsodiscussesthepossibilitythatearningsmanagementmightresultinsimilarempiricalpatternsandtriestodistinguishshelteringac-tivityfromearningsmanagement.
6ManzonandPlesko(2001)relatetheabsolutemagnitudeofthegaptopossibledetermi-nantsofdemandforshelteringactivity.
Theyfindthatdummyvariablesassociatedwiththepresenceofpretaxprofitsandnetoperatinglosseshelppredictthesizeofthegapinacross-sectionalregression.
7Thisdifferenceresultsfromthedeparturefrom"cleansurplus"accountingintheaccount-ingofstockoptions.
Inshort,taxbenefitsfromexerciseofESOsdonotflowthroughtheincomestatementbutaresimplytransferredtoadditionalpaid-incapital.
Asaconsequence,taxexpensesfromaccountingstatementswon'tallowforconsiderationofthedeductionsassociatedwithexercises.
Thereportingofthesetaxbenefitsassociatedwithemployeeexer-cisesiscontinuingtoevolve,andHanlonandShevlin(2001)arguethatvariedestimatesoftaxratesmakethemistakeofinferringtaxrateswithoutconsiderationfortheeffectsofthesetaxbenefits.
8SeeDechowandSkinner(2000),HealyandWahlen(1999),andSchipper(1989)forreviewarticlesofthisliterature.
TheDivergenceBetweenBookIncomeandTaxIncome1753.
TAXINCOME,SIMULATEDBOOKINCOME,ANDACTUALBOOKINCOMEConcernsoverincreasedshelteringactivitiesbycorporationsaretypicallyassociatedwithtrendsincorporatetaxreceiptsandtrendsinthegapbetweentaxincomeandbookincomeemployingfiguresextractedfromtaxreturns.
Figure1considerstheratiooffederalcorporatetaxreceiptstoallfederalon-budgettaxreceiptsfrom1971to2001,asreportedinOfficeofManagementandBudget(2002).
Thefiguredepictstheoveralldeclineinthatratio,fromnearly19.
7percentin1977to10.
2percentin2001.
Muchofthepatterninthisratiofortheinterveningyearscanbeexplainedbylargelegislativechangesandgeneraleconomicconditions.
Therecentdecline,beginningin1996,ismorepuzzlinggiventhecoinci-denteconomicexpansion.
Inparticular,thisratiofallsfrom15.
8percentin1996to13.
5percentin2000andthen,evenmoredramatically,to10.
2percentin2001.
TherelationshipbetweentaxandbookincomeasreportedinScheduleM-ldemonstratesasimilarlycurioustrend.
Thesefigureshavebeenem-ployedbytheTreasuryDepartmenttoemphasizetheincreasedincidenceandmagnitudeofcorporatetaxshelteractivity.
Figure2reproducesafigurefromTalisman(1999)thatisalsorelatedtothefigureemployedintheTreasuryDepartment'sstudyofcorporatetaxshelters(U.
S.
Treasury,FIGURE1.
CorporateTaxReceiptsasPercentageofTotalOn-BudgetReceipts,1971-2001251971197619811986199119962001Thegraphshowstheratioofcorporatetaxreceiptstoallon-budgetfederalreceipts,asreportedinOfficeofManagementandBudget(2002).
276DesaiB-BIIPQWDOSOONONXoX.
39xo120,pptooo3tax-ereaitaxeslessthassetsa^oJ§*-ooooOOocgg.
gX3>oSoT3cteen"a,!
CoTheDivergenceBetweenBookIncomeandTaxIncome1771999).
Thisgraphemploysdataforonlythosecorporationswithassetsgreaterthan$1billion.
Inferringamultipleofbookincomeovertaxin-comefromthatgraphillustratestheconcernthattax-shelteringactivityhasincreasedoverthatperiod.
Inparticular,theratioofbookincometotaxincomegrowsto1.
4from1.
0overfiveyears.
Alternatively,bookin-comeexceedstaxincomebyapproximately$120billion(in1992dollars)by1996.
Thisfigure,andthegapithighlights,hasservedasthemostimportantsourceofdataforthedebateoncorporatetaxshelters.
AspointedoutbyKies(1999a),however,thereareseveralalternativeexpla-nationsforthisgapthatmakesuchapictureinconclusive.
Tounderstandtheimplicationofthiswideninggapforthescopeoftax-shelteringactivity,itisusefultoconstructameasureofsimulatedbookincomethatincorporatesidentifiablesourcesofthatgap.
Inparticu-lar,discrepanciesbetweentaxandbookincomemaybeattributabletocausesunrelatedtotax-shelteringactivity.
TheevidenceinTable1andFigure3considersthedynamicsoftaxandbookincomeforfirmsthathaveassetsgreaterthan$250millionandisolatestheimpactstemmingfromthreepotentialsourcesofthatgap:(1)thedifferentialtreatmentofdepreciationexpenseontaxandbookforms,(2)reinvestedearningsabroad,and(3)deductionsassociatedwiththeexerciseofnonqualifiedstockoptions.
Beforeanalyzingthedata,thedistinctionsbetweenthetwosetsofdatashouldbeemphasized.
Figure2employsactualtaxreturndatafrom1991to1996toanalyzethegapbetweenactualtaxandbookincomeforcorporationsthathavemorethan$1billioninassetsandarenotS-corporations,RICs,orREITs.
Incontrast,Table1andFigure3ini-tiallysimulatethatgapforsimilarcorporations,butwithacutoffof$250millioninassetsovertheperiodfrom1982to2000.
Table1providesdata,incurrentdollars,forsimulatingbookincomeforthesecorporationsandbeginswithanestimateoftaxincomecompa-rabletothebaseintheTalismanfigure.
Toanalyzetheimpactofthediscrepancybetweendepreciationexpenseallowedfortaxpurposesandthatassociatedwithbookaccounting,Table1drawsonBureauofEco-nomicAnalysis(BEA)estimatesofthecapitalconsumptionallowance(CCA)adjustment,whichmeasuresthediscrepancybetweentaxmea-suresofdepreciationandeconomicdepreciation.
Thisaggregatemeasureofthediscrepancyisscaledbythatfractionofdepreciationexpensesforfirmswithover$250millioninassets.
Tothedegreethateconomicdepre-ciationisnotrepresentativeofdepreciationassociatedwithhistoriccostaccounting,andtothedegreethatthislinkbetweenaccountingandeco-nomicdepreciationhasbecomemoretenuousovertimegiventhechang-ingnatureofassets,theevidenceinTable1likelyunderstatestheimpactassociatedwithgapsbetweentaxandaccountingnotionsofdepreciation.
178DesaiTABLE1TaxIncomeandComponentsofSimulatedBookIncome,1982-2000*Year1982198319841985198619871988198919901991199219931994199519961997199819992000Taxincome85,223103,103130,569142,224136,194214,618306,665248,551227,723215,572251,587307,302357,936436,533485,645513,332457,575457,575457,575Excessdepreciation8,43719,34030,27643,55733,45731,91530,34123,75913,5427,0313,0142,8668,54513,09517,42723,66027,35637,21831,243Reinvestedearningsabroad4,57413,10716,80313,3979,24717,68713,44711,77520,48817,56815,73329,21223,34245,70846,03547,72831,62058,80684,111ProceedsfromoptionexercisesN/AN/AN/AN/AN/AN/AN/AN/AN/AN/A14,08615,36510,41617,64932,41242,62673,59874,832106,265Simulatedbookincome98,234135,550177,649199,179178,899264,220350,453284,085261,753240,171284,419354,744400,240512,985581,519627,346590,150628,431679,195*Alldollarfiguresareincurrentdollars.
Taxincomeistotalreceiptslesstotaldeductionsforfirmswithgreaterthan$250millioninassets,excludingRICs,REITs,andS-corporations.
Thetaxincomefiguresfor1999and2000arenotavailable,sothesimulatedbookincomefigureassumesthattaxincomestaysat1998levels.
ExcessdepreciationcorrespondstotheCCAadjustmentscaledfortherelativeuseofdepreciationbyfirmswithgreaterthan$250millioninassets.
Reinvestedearningsabroadarethediffer-encebetweenforeignearningsandrepatriationsscaledfortherelativelevelsofforeignincomeforfirmswithgreaterthan$250millioninassets.
Optionexercisesaretheproceedsfromexercises,asmorefullyreportedinTable2.
Simulatedbookincomeisthesumofthosefigures.
Similarly,Table1capturesthediscrepancybetweenincomeearnedbyU.
S.
corporationsoperatingabroadandtheincomethatisrepatriatedfromBEAdataoncapitalflows.
9Finally,Table1providesdatafromExecucompontheaggregatelevelofoptionexercisesbyemployees,whichisdiscussedingreaterdetailbelow.
Thesedataareavailableonlyforyearssubsequentto1992.
Taxincomefrom1999and2000isnotyetavailablebutispresumedtostayat1998levelstofacilitateabaselinecomparisonfor1999and2000.
Figure3relatesthefindingsofTable1inaformatcomparabletotheevidenceprovidedinFigure2bytranslatingthefiguresfromTable19Thesefiguresarescaledbytheannualfractionofforeigntaxcreditsattributabletofirmswithassetsmorethan$250million.
TheDivergenceBetweenBookIncomeandTaxIncome179COsuiCO"^fNiIi^^J^SOv—mov09\ON—soO—OS73>i*55.
S8*S&"tfflS>7-(inElacs2s^^.
s§1SsSi2.
SI-aS>.
2§I-8SIs:is.
gs-agigis**r"pj\—'toWIU,o73":[°^-Stui3^2u->Tfooo"gS23m^§TJ^&HH22-^OIy0cn"OHcn0)G"OHcnCDcncucn0)cn0)"EHcnOJ_C"HHcn(LI_gEHcnEHcnOJUtCUTheDivergenceBetweenBookIncomeandTaxIncome195FIGURE5.
BookIncomeVersusTaxableIncome,SplineEstimatesbyThree-YearIntervalswithIndustryFixedEffects,1992-2000—1992-1994456ln(taxableincome)1995-1997101998-2000ThethreelinesinthefigureareconstructedusingthecoefficientsofsplineregressionsreportedinTable5,whereindustryfixedeffectsareemployedforthegroupsofyears1992-1994,1995-1997,and1998-2000.
FIGURE6.
BookIncomeVersusTaxableIncome,SplineEstimatesbyYearwithIndustryFixedEffects,1992-2000123456ln(taxableincome)199219931994199719981999—o1995-2000101996ThelinesinthefigureareconstructedusingthecoefficientsofsplineregressionsreportedinTable6ofDesai(2002),whereindustryfixedeffectsareemployedforeachyearfrom1992to2000.
196DesaiGiventheindustry-specificityofdiscretionaryaccrualopportunities,itisusefultoemployindustryfixedeffectsinthesepiecewiselinearregres-sionsinanefforttoisolateatax-shelteringexplanationforthisphenome-nonfromanearningsmanagementperspective.
TheresultspresentedinTable5,withtheinclusionofindustryfixedeffects,demonstratethatwithin-industryvariationprovidesevenstrongerevidenceforaweaken-ingoftherelationshipbetweenbookincomeandtaxincomeatlowlevelsoftaxincome.
ThecoefficientsfromthethreespecificationsemployingindustryfixedeffectsareusedtoconstructthelinesinFigure5.
Thisfigureprovidestheflatteningoftherelationshipbetweenbookincomeandtaxincome,aspredictedinthemodelofcostlyshelteringoverthecourseofthe1990s.
Thissameexerciseisrepeatedbyyearwithindustryfixedeffects,andtheresultsarepresentedingraphicalforminFigure6.
Fig-ure6providessimilarresultsbecausethecoefficientsonthefirstsplinearepositiveandsignificantinearlyyearsofthesampleandbecausetheyapproach0by2000.
AsreportedinDesai(2002),theexplanatorypoweroftheregressionsthatunderliesFigure6declinesoverthesampleperiod,furtheremphasizingthereducedlinkbetweenbookincomeandtaxincome.
Thisregressionevidenceisdifficulttoreconcilewithalternativeratio-nalesforthedisparitybetweentaxincomeandbookincome.
Forexample,ifthedifferentialmeasurementofdepreciationorofreinvestedearningsabroadweretoaccountwhollyforthischangedpattern,thisdifferentialmeasurementwouldhavetohavebecomemoreconcentratedamonglow-tax-incomefirmsduringthe1990s.
Whilethesepossibilitiesappearun-likely,itispossiblethattheaccountingofoptionexercises,iffirmsfollowcleansurplusaccounting,wouldhavereducedtaxincome,asmeasuredbyaccountingstatements,foryounggrowthfirmswithlowtaxincome.
Controllingseparatelyforthelevelsofoptionactivityintheseregressionsdemonstratesthatoptionactivitycannotexplainthisdecouplingofbookincomeandtaxincomeatlowlevelsoftaxincome.
Toensurethatresultsarenotdisproportionatelyreflectingthedifferentbehaviorofsmallorlargefirms,appendixtable2andappendixfigureslaandlbofDesai(2002)attemptthesamepiecewiselinearregressionsprovidedinTable5bydividingthesampleatthemedianlevelofsales.
Thesamepatternsappeartohold.
hiadditiontousingindustryfixedeffectstoseparateearningsmanage-mentexplanationsfromtax-shelteringexplanations,itispossibletocon-sidertherelativevariabilityofbookincomeandtaxincomeinanefforttosegregateactiveearningsmanagersfromnonearningsmanagers.
26To26Suchadistinctionresemblesotherstudiesthatattempttoisolateearningssmoothersbytherelativeabsenceofvariabilityinreportedearnings,asinMyersandSkinner(2001),whostudyfirmswithconsistentincreasesinreportedearnings.
TheDivergenceBetweenBookIncomeandTaxIncome197boingsOu-*ON,AON.
"°fc3-^i-JSr-HCOOIboICOCOONOINoooincoNOi-H^oNNbcMcoLNONOoocooooinooNoooo0^TticooNcoinco^cMoOi—iGOrHOOooLn10in*CMCMcoINCMOOOOOOOoooooOOOOOrHOOONOOCMNOON"rtlLNrHinCMi-HlN'HHTtlOOOi-H-^T}01mlineNOlinea,(X)LNlineen0001"OHenONlineen*oi—ilineen"01ff!
zOIc^5gS$eggniSU,^H5b13ggog'uOHCJOHoOHOCUTjouc"o-UingS^f"""ou§.
aacatSH'SfuQJJBCnsoCtnM)vaS°-2-OJo5J.
5g"43TJaT3T3^nsto"5MHiSOJ°goi'Cg_^fill198Desaidoso,ratiosofthestandarddeviationsofbookincometotaxincomearecalculatedforallthefirmsinthesample.
Thisratiohasamedianof0.
91,andfirmswitharatiobelowthemedianareconsideredearnings-smoothingfirmsandfirmswitharatioabovethemedianareconsiderednonsmoothers.
Ifearningmanagementwereaprimarydriverofthebreakdownoftherelationshipbetweenbookincomeandtaxincomeatlowlevelsoftaxincome,thenfirmswithdifferentpropensitiestosmoothbookincomeshouldexhibitadifferentrelationshipbetweentaxincomeandbookincomeatlowlevelsoftaxincomeoverthecourseofthedecade.
TheevidenceprovidedinTable6andFigures7and8suggestotherwisebecausebothearningssmoothersandnon-earningssmoothersexhibitsimilarunderlyingtrendsoverthecourseofthedecadewithrespecttoaflatteningofthebookincome-taxincomerelationshipatlowlevelsoftaxincome.
TheregressionframeworkexploredinTables5and6andFigures5to8providesevidencethatisconsistentwiththemodeloftaxshelteringFIGURE7.
BookIncomeVersusTaxableIncome,Earnings-SmoothingFirms,SplineEstimatesbyThree-YearIntervalswithIndustryFixedEffects,1992-200010-345678910ln(taxableincome)—1992-19941995-1997-X-1998-2000ThethreelinesinthefigureareconstructedusingthecoefficientsofsplineregressionsreportedinTable6forthegroupsofyears1992-1994,1995-1997,and1998-2000.
Thefigureemploysregressionsfortwosamplesoffirmsthataredividedatthemedianratioofthestandarddeviationofbookincometothestandarddeviationoftaxableincome.
Firmswithratiosabovethemedianratioaretermednonearnings-smoothingfirms,andthosewithratiosbelowthemedianratioaretermedearnings-smoothingfirms.
TheDivergenceBetweenBookIncomeandTaxIncome199FIGURE8.
BookIncomeVersusTaxableIncome,Nonearnings-SmoothingFirms,SplineEstimatesbyThree-YearIntervalswithIndustryFixedEffects,1992-20000+-012^—1992-19943456ln(taxableincome)1995-19977->8e—1998-2000ThethreelinesinthefigureareconstructedusingthecoefficientofsplineregressionsreportedinTable6forthegroupsofyears1992-1994,1995-1997,and1998-2000.
Thefigureemploysregressionsfortwosamplesoffirmsthataredividedatthemedianratioofthestandarddeviationofbookincometothestandarddeviationoftaxableincome.
Firmswithratiosabovethemedianratioaretermednonearnings-smoothingfirms,andthosewithratiosbelowthemedianratioaretermedearnings-smoothingfirms.
providedinsection5.
2.
Theimplicationoftheseresultsisthatshelteringbecameconsiderablylesscostly,eitherthroughloweredprobabilitiesofdetectionorperceivedlowerpenalties,andthatfirmsbecamemoreag-gressiveduringthe1990s.
Thealternativeexplanationofasecularin-creaseinearningsmanagementisdifficulttoreconcilewiththedataprovidedinthepaper.
First,earningsmanagementtheoriestypicallydonotallowforsuchlong-runintertemporalshiftingofincome,aswouldberequiredtoexplaintheaggregatetrendsinFigure3.
Second,ifearningsmanagementopportunitiesareparticularlyassociatedwithdiscretionaryaccrualopportunitiesinsomeindustries,thenwithin-industryvariationshouldhavereduced,ratherthanincreased,evidenceofthisunderlyingbehaviorinthemicroevidence.
Finally,ifincreasedearningsmanage-mentwerethecauseoftheflatteningoftherelationshipbetweenbookincomeandtaxincomeatlowlevelsoftaxincome,simpledistinctionsintherelativevariabilityofbookincomeandtaxincomeshouldhavedemonstratedmorepronounceddifferencesinthisbreakdown.
Whileitisnotpossibletoruleoutearningsmanagementdefinitively—particularly200Desaifraudulentbookreportingofincome—asasourceofsomeoftheaggre-gatephenomenaobservedinFigure3,thismicroanalysissuggeststhatthedistinctivewayinwhichtherelationshipbetweenbookincomeandtaxincomehasdeterioratedoverthedecadeisconsistentwithincreasedlevelsofcostlysheltering.
6.
CONCLUSIONThispaperattemptstoilluminatethedebateontaxshelteringbydisentan-glingvariedexplanationsforthebreakdownintherelationshipbetweentaxincomeandbookincomeoverthelastdecade.
First,thepaperdemon-stratesthatidentifiablefactorsthatonceaccountedforthedifferencebe-tweentaxincomeandbookincomefailtoaccountforthelargegapbetweenbookincomeandtaxincomeinmorerecentyears.
Inparticular,by1998,morethanhalfofthedifferencebetweentaxincomeandbookincome—approximately$154.
4billion,or33.
7percentoftaxincome—cannotbeaccountedforbythesehistoricallyrelevantmeasuresofthediscrepancybetweentaxincomeandbookincome.
Second,thepaperdemonstratesthatwithinthoseidentifiablefactors,changedpatternsinemployeecompensationarecreatingthelargestdisparitybetweenbookincomeandtaxincome.
By1998,differenttreatmentofemployeeoptionsaccountedfornearlythreetimesasmuchofthedifferencebetweenbookincomeandtaxincomeasdiddistinctdefinitionsofdepreciation.
Byexaminingthissamedisparitybetweenbookincomeandtaxincomeusingaccountingdata,thispaperdemonstratesthatthisbreakdownhasbeenparticularlypronouncedatlowlevelsoftaxincomeandhasdegen-eratedprogressivelyoverthedecade.
Theseresultsareconsistentwiththemodelofcostlytaxshelteringpresentedinthepaperanddecreasedcostsofshelteringoverthedecadeleadingtogreatergapsbetweentaxincomeandbookincome.
Teststocheckiftheseresultsaredrivenbyincreasedlevelsofearningssmoothingdonotappeartodiminishthere-sultsthatsuggestthatincreasedtaxshelteringisresponsibleforthedis-tinctivebreakdownintherelationshipbetweenbookincomeandtaxincome.
ThelargediscrepancyreportedbetweensimulatedbookincomeandactualbookincomefromFigure3,alongwiththeregressionevidence,suggeststhateffortsbyfirmstocircumventtaxpaymentsarebecomingmoresignificant,cheapertoimplement,andhardertodetect.
Thesedevel-opmentsprovideyetanotherreasontoreevaluatethemannerinwhichcorporateearningsaretaxedbecausetheunderlyingdevelopmentsdriv-ingthesephenomena—includingincreasedaccesstoglobalopportunitiesandtherapiddevelopmentoffinancialinnovations—areunlikelytode-TheDivergenceBetweenBookIncomeandTaxIncome201clineinimportanceinthenearfuture.
Similarly,thesetrendsillustratetheneedtorevisittherationalesfordistinctivebookandtaxreporting,andthedegreeandmannerinwhichfirmsareforcedtoreconcilebookincomeandtaxincome.
APPENDIXInordertomotivatetheempiricaltestsofincreasedtaxsheltering,itisusefultospecifyamodelofcostlytaxsheltering.
LetIequaltrueeconomicincome,Bequalreportedbookincome,andTequalreportedtaxincome.
Bookincomeispresumedtocorrespondtotrueeconomicincomewithnoise,sothat:B=Ieuwhereu~N(0,a2)(1)Tospecifywhattaxincomewouldbe,letsequalshelteredincome,whichisassociatedwithcostsy(s),whichinturnarecharacterizedbyy'(s)>0andy"(s)>0.
Inthissetting,thecostsofshelteringcorrespondonlytotheamountofincomeshelteredandnottheamountoftrueeco-nomicincome.
Inotherwords,largerfirmshavenoadvantageinshelter-ingincomerelativetosmallerfirms.
Asaconsequence,taxincomeisgivenby:T=I-s-y(s)(2)andthetaxrateisafunctionofthistaxableincome,asrepresentedbyx[I-s-y(s)].
Firmschoosetheamountofincometoshelterbysolving:max[I-s-y(s)]{l-T[I-s-y(s)]}+s(3)swhichyieldsthefirst-ordercondition:-[1+y'(s)][l-T(-)]+[I-s-y(s)][l+y'(s)]T'O+1=0(4)Toanalyzethisproblemfurther,itisusefultoconsidertworegionsofthetaxschedule:onewherethereisprogressivityandonewherethere202Desaiisnoprogressivity.
Atsufficientlyhighlevelsofincome,thetaxratewillexhibitnoprogressivity;T'Q=0,andequation(4)willcollapseto:[1+y'(s)][l-T0]=1(5)Atsuchlevelsofincome,levelsofshelteringwillnolongervarywithlevelsofincomebecause:dlAsaconsequence,itispossibletorewritetheleveloftaxincomeinequa-tion(2)as:T=I-s-y(s)=I-k(6)Takinglogarithmsofbothsidesofequations(6)and(1)andemployingafirst-orderTaylorapproximation,itispossibletorewriteequation(1)as:ln(B)=ln(I)+-+u(7)whichinturnbecomestheestimatingequationintheempiricalanalysisthatfollows.
Whenlevelsofincomearesuchthatx'(-)^0,thefirst-orderconditioninequation(4)canberewrittenasanexpressionfortaxincome:(1-T)-Y(S)](8)Tomakethisequationtractable,itisusefultotransformitinto:1+y'(s)(9)whereF(x)=[T'(-)X+x(-)—1],andtodefinetheinversefunctionasG(z)=F~l(z).
Thisspecificationallowstaxincometoberewrittenas:TheDivergenceBetweenBookIncomeandTaxIncome203T=G—(10)\l+Y(s)JThecurvatureofG(-)willbesmallwhenthecurvatureofF(-)ishigh,soitbecomesclearthatatlowlevelsofincome=0.
dlWithrespecttotheestimatingequationprovidedinequation(7),thismodelsuggeststhattherelationshipbetweenbookincomeandtaxin-comewillbedistinctiveatlowlevelsoftaxincome,necessitatingapiecewiselinearmodeltoestimateequation(7).
Morespecifically,themodelsuggeststhattaxshelteringwillbereflectedbyarelativelyflatterrelationshipbetweenbookincomeandtaxincomeatlowlevelsoftaxincome.
Correspondingly,increasedlevelsoftaxshelteringwillbere-flectedinaflatteningoftherelationshipbetweentaxincomeandbookincome.
REFERENCESAdams,J.
(2001).
"APensionAccountingPrimer.
"CSFBEquityResearchReport.
June.
Auerbach,A.
,andJ.
Poterba(1987).
"WhyHaveCorporateTaxRevenuesDe-clined"InTaxPolicyandtheEconomy,vol.
1,L.
Summers(ed.
).
Cambridge,MA:MITPress.
Bankman,J.
(1999).
"TheNewMarketinCorporateTaxShelters.
"TaxNotesToday83:1775-1795.
Barnea,A.
,J.
Ronen,andS.
Sadan(1975).
"TheImplementationofAccountingObjectives—AnApplicationtoExtraordinaryItems.
"TheAccountingReview50:58-68.
BearStearns(2000).
"EmployeeStockOptions.
"AccountingIssues(July6).
Brenner,M.
,R.
Sundaram,andD.
Yermack(2000).
"AlteringtheTermsofExecu-tiveStockOptions.
"JournalofTinancialEconomics57:103-128.
Burgstahler,D.
,andI.
Dichev(1997).
"EarningsManagementtoAvoidEarningsDecreasesandLosses.
"JournalofAccountingandEconomics24(1):99-126.
Callihan,D.
(1994).
"CorporateEffectiveTaxRates:ASynthesisoftheLiterature.
"JournalofAccountingLiterature13:1-43.
27Moreprecisely,thisequationrequirestheadditionalassumptionthat%"{)=0andthaty"(-)^T'Obecausef(s)f1dly"(s)+2T'Q[1+Y'(S)]3204DesaiCore,J.
,andW.
Guay(2001).
"StockOptionPlansforNon-ExecutiveEmployees.
"JournalofFinancialEconomics61(2):253-287.
Dechow,P.
,andD.
Skinner(2000).
"EarningsManagement:ReconcilingtheViewsofAccountingAcademics,Practitioners,andRegulators.
"AccountingHorizons14(2):235-250.
Degeorge,F.
,J.
Patel,andR.
Zeckhauser(1998).
"EarningsManagementtoExceedThresholds.
"JournalofBusiness72(l):l-33.
Desai,M.
(2002).
"TheCorporateProfitBase,TaxShelteringActivity,andtheChangingNatureofEmployeeCompensation.
"NBERWorkingPaperno.
8866.
March.
Desai,M.
,C.
Foley,andJ.
HinesJr.
(2001).
"RepatriationTaxesandDividendDistortions.
"NationalTaxJournal54:829-851.
Desai,M.
,C.
Foley,andJ.
HinesJr.
(2003).
"ChainsofOwnership,RegionalTaxCompetition,andForeignDirectInvestment.
"InForeignDirectInvestmentintheRealandFinancialSectorofIndustrialCountries,HeinzHerrmannandRobertLip-sey(eds.
).
Heidelberg,Germany:Springer-Verlag.
Desai,M.
,andJ.
HinesJr.
(2002).
"ExpectationsandExpatriations:TracingtheCausesandConsequencesofCorporateInversions.
"NationalTaxJournal55(3):409-440.
Dworin,L.
(1985).
"OnEstimatingCorporateTaxLiabilitiesfromFinancialState-ments.
"TaxNotesToday85:965-971.
Dye,R.
(1988).
"EarningsManagementinanOverlappingGenerationsModel.
"JournalofAccountingResearch26(2):195-235.
Engel,E.
,M.
Erickson,andE.
Maydew(1999).
"Debt-EquityHybridSecurities.
"JournalofAccountingResearch37(2):249-274.
Fudenberg,D.
,andJ.
Tirole(1995).
"ATheoryofIncomeandDividendSmoothingBasedonIncumbencyRents.
"JournalofPoliticalEconomy103(l):75-93.
Goel,A.
,andA.
Thakor(2003).
"WhyDoFirmsSmoothEarnings"ForthcominginJournalofBusiness76(1).
Goolsbee,A.
(2000a).
"WhatHappensWhenYouTaxtheRichEvidencefromExecutiveCompensation.
"JournalofPoliticalEconomy108(2):352-378.
(2000b).
"Taxes,High-IncomeExecutives,andthePerilsofRevenueEsti-mationintheNewEconomy.
"TheAmericanEconomicReview90(2):271-275.
Hall,B.
,andJ.
Liebman(2000).
"TheTaxationofExecutiveCompensation.
"InTaxPolicyandtheEconomy,vol.
14,J.
Poterba(ed.
).
Cambridge,MA:MITPress.
Hanlon,M.
,andT.
Shevlin(March2002).
"AccountingforTaxBenefitsofEm-ployeeStockOptionsandImplicationsforResearch.
"PublishedinAccountingHorizons16:1-16.
Healy,P.
,andJ.
Wahlen(1999).
"AReviewoftheEarningsManagementLitera-tureandItsImplicationsforStandardSetting.
"AccountingHorizons13(4):365-383.
Huddart,S.
,andM.
Lang(1996).
"EmployeeStockOptionExercises:AnEmpiricalAnalysis.
"JournalofAccountingandEconomics21:5-43.
JointCommitteeonTaxation(2000).
"TestimonyoftheStaffoftheJointCommit-teeonTaxationConcerningInterestandPenaltiesandCorporateTaxSheltersBeforetheSenateCommitteeonFinance.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
March.
Jones,J.
(1991).
"EarningsManagementDuringImportReliefInvestigations.
"JournalofAccountingResearch29:193-228.
Kies,K.
(1999a).
"StatementofKennethKies:TestimonyBeforetheHouseCom-TheDivergenceBetweenBookIncomeandTaxIncome205mitteeonWaysandMeansHearingonPenaltyProvisioninPresident'sFiscalYear2000Budget.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
March.
(1999b).
"ACriticalLookattheAdministration's'CorporateTaxShelter'Proposals.
"TaxNotesToday83:1463-1486.
(1999c).
"StatementofKennethKies:TestimonyBeforetheHouseCommit-teeonWaysandMeansHearingonCorporateTaxShelters.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
November.
(2000).
"StatementofKennethKies:TestimonyBeforetheSenateFinanceCommitteeHearingonPenaltyandInterestProvisionsintheInternalRevenueCode.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
March.
Kinney,M.
,andE.
Swanson(1993).
"TheAccuracyandAdequacyofTaxDatainCompustat.
"JournalofAmericanTaxationAssociation15:121-135.
Mackie,].
,Ill(2000).
"TheComebackoftheCorporateIncomeTax.
"InProceedingsoftheNinety-SecondAnnualConferenceonTaxation.
Washington,D.
C.
:NationalTaxAssociation,93-102.
Manzon,G.
,Jr.
,andG.
Plesko(2001).
"TheRelationBetweenFinancialandTaxReportingMeasuresofIncome.
"TaxLawReview55(2):175-214.
Mclntyre,R.
(2000).
"CorporateIncomeTaxesinthe1990s.
"Washington,D.
C.
:InstituteonTaxationandEconomicPolicy.
October.
Miller,G.
,andD.
Skinner(1998).
"DeterminantsoftheValuationAllowancesforDeferredTaxAssetsUnderSFASNo.
109.
"TheAccountingReview73(2):213-233.
Mills,L.
,andK.
Newberry(2001).
"TheInfluenceofTaxandNon-TaxCostsonBook-TaxReportingDifferences:PublicandPrivateFirms.
"TheJournaloftheAmericanTaxationAssociation23(1).
Murphy,K.
(1999).
"ExecutiveCompensation.
"InHandbookofLaborEconomics,vol.
Ill,OrleyAshenfelterandDavidCard(eds.
).
NewYork:North-Holland.
Myers,L.
,andD.
Skinner(2001).
"EarningsMomentumandEarningsManage-ment.
"UniversityofMichiganWorkingPaper.
OfficeofManagementandBudget(2002).
"HistoricalTables:FiscalYear2003BudgetoftheU.
S.
Government.
"Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
Omer,T.
,K.
Molloy,andD.
Ziebart(1991).
"MeasurementofEffectiveCorporateTaxRatesUsingFinancialStatementInformation.
"JournaloftheAmericanTaxa-tionAssociation13:57-72.
Petrick,K.
(2001).
"ComparingNIPAProfitswithS&P500Profits.
"SurveyofCur-rentBusiness(April):16-20.
Plesko,G.
(2000).
"Book-TaxDifferencesandtheMeasurementofCorporateIn-come.
"InProceedingsoftheNinety-SecondAnnualConferenceonTaxation.
Wash-ington,D.
C.
:NationalTaxAssociation.
(2002).
"ReconcilingCorporationBookandTaxNetIncome,TaxYears1996-1998.
"SOIBulletin.
Washington,D.
C.
:U.
S.
GovernmentPrintingOffice.
(2003).
"AnEvaluationofAlternativeMeasuresofCorporateTaxRates.
"ForthcominginJournalofAccountingandEconomics,February2003.
Porcano,T.
(1986).
"CorporateTaxRates:Progressive,Proportional,orRegres-sive.
"JournaloftheAmericanTaxationAssociation7:17-31.
Schipper,K.
(1989).
"Commentary:EarningsManagement.
"AccountingHorizons3(4):91-102.
Shevlin,T.
(1987).
"TaxesandOff-BalanceSheetFinancing:ResearchandDevel-opmentLimitedPartnerships.
"TheAccountingReview62(3):480-509.
206DesaiStickney,C,andV.
McGee(1982).
"EffectiveCorporateTaxRates—TheEffectofSize,CapitalIntensity,Leverage,andOtherFactors.
"JournalofAccountingandPublicPolicy1:125-152.
Sullivan,M.
(1999a).
"ShelterFalloutCorporateTaxesDown,ProfitsUp.
"TaxNotesToday84:653-657.
(1999b).
"ARevenueEstimateforCorporateTaxShelters.
"TaxNotesToday85:981-983.
(2000a).
"Lobbyist'sFiguresFlawed,DataIndicateCorporateShortfalls.
"TaxNotesToday86:309-313.
(2000b).
"LettheGoodTimesRoll:OptionsandTax-FreeProfits.
"TaxNotesToday87:1185-1191.
Talisman,J.
(1999).
"StatementofJonathanTalisman:TestimonyBeforetheHouseCommitteeonWaysandMeansHearingonCorporateTaxShelters.
"Washing-ton,D.
C.
:U.
S.
GovernmentPrintingOffice.
November.
Trueman,D.
,andS.
Titman(1988).
"AnExplanationforAccountingIncomeSmoothing.
"JournalofAccountingResearch26:127-139.
U.
S.
DepartmentoftheTreasury(1999).
"TheProblemofCorporateTaxShelters:Discussion,AnalysisandLegislativeProposals.
"Washington,D.
C.
:U.
S.
Gov-ernmentPrintingOffice.
Zimmerman,J.
(1983).
"TaxesandFirmSize.
"JournalofAccountingandEconomics5:119-149.
EditedbyJamesM.
PoterbaNationalBureauofEconomicResearch0-262-66138-OfrelatedinterestTaxPolicyandtheEconomy,Volume16EditedbyJamesM.
PoterbaArticlesbyRobertMoffitt;DavidFiglio;JonathanGruber;BrigitteMadrianandDavidLaibson;RonaldLee;CharlesMcLureTaxPolicyandtheEconomy,Volume15EditedbyJamesM.
PoterbaArticlesbyMihirDesaiandJamesHines,Jr.
;RobertCarroll,DouglasHoltz-Eakin,MarkRider,andHarveyRosen;AndrewMitrusiandJamesM.
Poterba;DavidEllwoodandJeffreyLiebman;MartinFeldsteinandElenaRanguelova;MarkMcClellanTheMITPressMassachusettsInstituteofTechnologyCambridge,Massachusetts02142http://mitpress.
mit.
edu
- requirementscncn相关文档
- 食品cncn
- 基地cncn
- 苍南县龙港镇林陈至四大(龙港段)联网公路工程Ⅰ标段
- 指数cncn
- classescncn
- I20cncn
DiyVM：2G内存/50G硬盘/元起线路香港vps带宽CN2线路,香港VPS五折月付50元起
DiyVM是一家低调国人VPS主机商，成立于2009年，提供的产品包括VPS主机和独立服务器租用等，数据中心包括香港沙田、美国洛杉矶、日本大阪等，VPS主机基于XEN架构，均为国内直连线路，主机支持异地备份与自定义镜像，可提供内网IP。最近，商家对香港机房VPS提供5折优惠码，最低2GB内存起优惠后仅需50元/月。下面就以香港机房为例，分享几款VPS主机配置信息。CPU：2cores内存：2GB硬...
RAKsmart 年中活动 独立服务器限时$30秒杀 VPS主机低至$1.99
RAKsmart 虽然是美国主机商，但是商家的主要客户群还是在我们国内，于是我们可以看到每次的国内节日促销活动期间商家也会发布促销。包括这次年中大促活动，RAKsmart商家也有发布为期两个月的年终活动，其中有商家擅长的独立服务器和便宜VPS主机。服务器包括站群服务器、特价服务器、高达10G带宽不限制流量的美国服务器。商家优惠活动，可以看到对应商品的优惠，同时也可以使用 优惠码 RAKBL9 同时...
GigsGigsCloud（$26/年）KVM-1GB/15G SSD/2TB/洛杉矶机房
GigsGigsCloud新上了洛杉矶机房国际版线路VPS，基于KVM架构，采用SSD硬盘，年付最低26美元起。这是一家成立于2015年的马来西亚主机商，提供VPS主机和独立服务器租用，数据中心包括美国洛杉矶、中国香港、新加坡、马来西亚和日本等。商家VPS主机基于KVM架构，所选均为国内直连或者优化线路，比如洛杉矶机房有CN2 GIA、AS9929或者高防线路等。下面列出这款年付VPS主机配置信息...
cncn com为你推荐
-
回收站在哪回收站去哪里了？google竞价排名谷歌是靠什么关键词排名的？谷歌竞价有什么特点？有什么优势？iphone5解锁iphone5密码忘了怎么解锁照片转手绘照片弄成手绘一样的那个软件到底叫什么，能不能告诉啊？安卓应用平台安卓系统支持的软件并不是那么多，为什么这么多人推崇？小米手柄手机用小米手柄能玩什么游戏声母是什么声母.韵母有哪些网站推广外链如何做网站推广 ，外链推广的方向在哪里？av终结者专杀工具av终结者木马病毒专杀工具没用ncsetting为什么安装NCSettings没有反应啊，那个开关是默认开的啊，可通知栏没有这个啊？