The TCJA and Domestic Corporate Tax Rates

Finance and Economics Discussion Series Federal Reserve Board, Washington, D.C. ISSN 1936-2854 (Print) ISSN 2767-3898 (Online) The TCJA and Domestic Corporate Tax Rates Christine L. Dobridge, Patrick Kennedy, Paul Landefeld, Jacob Mortenson 2023-078 Please cite this paper as: Dobridge, Christine L., Patrick Kennedy, Paul Landefeld, and Jacob Mortenson (2023). “The TCJA and Domestic Corporate Tax Rates,” Finance and Economics Discussion Series 2023-078. Washington: Board of Governors of the Federal Reserve System, https://doi.org/10.17016/FEDS.2023.078. NOTE: Staff working papers in the Finance and Economics Discussion Series (FEDS) are preliminary materials circulated to stimulate discussion and critical comment. The analysis and conclusions set forth are those of the authors and do not indicate concurrence by other members of the research staff or the Board of Governors. References in publications to the Finance and Economics Discussion Series (other than acknowledgement) should be cleared with the author(s) to protect the tentative character of these papers.

The TCJA and Domestic Corporate Tax Rates We study changes in tax positions for U.S. C corporations following passage of the2017taxlegislationcommonlyknownastheTaxCutsandJobsAct(TCJA). While existing research has focused primarily on publicly traded companies, data limitations have prevented more holistic analyses of the corporate sector. Using a representative sample of U.S. corporate tax returns, we highlight how trends in effective tax rates (ETRs) and exposure to the legislation’s main provisions varied for public, private, multinational, domestic, and large versus small firms. We document several novel facts, including that ETRs increased on average for privately held, domestic firms and for firms in the bottom 90% of the firm sales distribution after TCJA. In contrast, public, multinational, and large firms saw substantial ETR cuts on average. We find that firms’ pre-TCJA exposure to changes in the corporate tax rate and treatment of net operating losses have the strongest correlation with post-TCJA ETR changes. Overall, the analysis underscores the divergent impacts of TCJA on different firm types and illuminates the economic scope and relative significance of TCJA’s myriad provisions. Keywords: Corporatetaxes,TaxCutsandJobsAct,taxreform JELClassification: H2,H25 ChristineL.Dobridge PatrickKennedy BoardofGovernorsoftheFederal NationalBureauofEconomicResearch& JointCommitteeonTaxation, ReserveSystem U.S.Congress PaulLandefeld◇ JacobMortenson JointCommitteeonTaxation, JointCommitteeonTaxation, U.S.Congress U.S.Congress ◇Correspondingauthor: Paul.Landefeld@jct.gov,502FordHouseOfficeBuilding,Washington,DC20515. This researchembodiesworkundertakenforthestaffoftheJointCommitteeonTaxation,butasmembersofbothparties andbothhousesofCongresscomprisetheJointCommitteeonTaxation,thisworkshouldnotbeconstruedtorepresent thepositionofanymemberoftheCommittee. Theviewsandopinionsexpressedherearetheauthors’own. Theyare not necessarily those of the Board of Governors of the Federal Reserve System, its members or its staff. Co-author ChristineDobridgedidnothaveaccesstotaxpayer-identifiabledatawhileworkingonthisproject. WethankMatthew Holt,JaneSong,ErinTowery,andparticipantsatthe2021NationalTaxAssociationSpringSymposiumandthe2021 NationalTaxAssociationAnnualConferenceforhelpfulcomments.

In 2017 the United States Congress passed legislation commonly known as the Tax Cuts and Jobs Act (TCJA), which introduced the most sweeping changes to American business taxation since the Tax Reform Act of 1986.1 In this paper, we answer several fundamental descriptive questionsaboutfirms’taxpositionsbeforeandafterthelaw’senactment: Howdidtheeffectivetax rates (ETRs) paid by C corporations change after TCJA? How do changes in ETRs vary with firm characteristics, such as firm size and whether a firm is public, private, domestic, or multinational? AndwhichofTCJA’smanyprovisionsappearmostrelevantinexplainingchangesinfirms’ETRs? Despite the importance of these questions for understanding and analyzing policy, existing research has not been able to comprehensively answer them, primarily due to data limitations. While prior empirical studies have focused on publicly traded firms, research suggests that smaller private domestic firms and startups are engines of innovation and growth, especially in economically important sectors such as technology and health (Decker et al., 2014). These types of firms are concentrated in private domestic firms, which account for the vast majority of C corporations and approximately one-third of C corporation employment. Among multinational firms, private C corporations outnumber their public counterparts five-to-one, and comprise approximately 18 percent of total C corporation employment. Understanding trends in both public and private firms is thus critically important for analyzing TCJA’s effects on the broader U.S. economy. Moreover, even among public firms, limitations in commonly used databases have prevented researchers from documenting key information such as, for example, the share of firms subjecttothecorporatealternativeminimumtaxpriortoitspost-TCJArepeal. We fill these gaps in existing research using a representative sample of de-identified corporate tax returns from 2010-2019. The rolling panel from the Internal Revenue Service’s Statistics of Income (SOI) files includes both public and private firms, and both domestic and multinational firms, allowing us to provide a holistic picture of the changing landscape of firms’ tax positions beforeandafterTCJA. Theanalysisproceedsinthreeparts. First,wedocumenttimeseriestrendsinaveragedomestic cash ETRs for different groups of firms, highlighting especially the divergence in trends for public/private, domestic/multinational, and small/large firms. Second, we show how the full distribution of firms’ ETRs shifted before and after TCJA. Third, we document basic evidence on the share of firms likely to have been affected by TCJA’s various provisions (which we call “exposuremeasures”),andshowhowtheseexposuremeasurescorrelatewithfirms’ETRchanges afterTCJA.Wenotethatthedescriptiveanalysisisnotintendedtoprovideevidenceonthecausal effects of TCJA’s individual provisions; rather, our aim is to highlight previously undocumented patterns, to provide clear evidence on the relative scope and economic significance of TCJA’s variousprovisions,andtodiscussimplicationsforpolicyanalysisandfutureresearch. 1Public Law 115-97 is titled ”An act to provide for reconciliation pursuant to titles II and V of the concurrent resolutiononthebudgetforfiscalyear2018.” 1

From the analysis, we document three descriptive facts. First, we show that ETRs declined sharply for publicly traded firms and for privately owned multinationals. These firms pay the lion’s share of federal corporate income taxes and account for approximately 69% of C corporationemployment. ThisresultisbroadlyconsistentwithHenryandSansing(2018);Dyreng et al. (Forthcoming); Wagner, Zeckhauser, and Ziegler (2020); however, our analysis paints a more nuanced picture relative to existing research due to our focus on domestic cash ETRs for multinationalfirms,asopposedtoglobalETRs. The second fact—which, from our perspective, appears to be less commonly appreciated in public and academic discourse—is that, in contrast to publicly traded and multinational firms, average ETRs increased for private domestic firms. While private domestic firms do not pay a large share of federal corporate income taxes, they do account for the vast majority of corporate firmsand31%ofCcorporationemployment. AnanalysisbasedonmatchingfirmsintheSOIpanel between 2016 and 2019 shows that a large majority (approximately 81%) of U.S. C corporations withpositivebookincome,representingasizeablefractionofU.S.salesandemployment,faceda corporateincometaxincreasefollowingTCJA. Third, when assessing the relative importance of TCJA’s various corporate provisions, we find that pre-TCJA exposure to the marginal corporate income tax rate changes (i.e., the increase in rates at the bottom of the tax schedule, and the decrease at the top) and the net operating loss (NOL) restriction are most strongly correlated with changes in firms’ ETRs after TCJA. By contrast, exposure to the alternative minimum tax, interest limitation, and multinational provisions was moderately correlated with changes in firms’ ETRs, while exposure to bonus depreciation-related changes and repeal of the domestic production activities deduction was only very weakly associated with changes in firms’ ETRs. We also discuss how firms’ behavioral responsestotheseprovisionsmaygeneratesecond-orderchangesinETRs,whicharealsocaptured inourestimates. These findings contribute to researchers’ understanding of this historically large corporate tax reform along two primary dimensions. First, we use a dataset that is representative of public and private firms in the United States. Virtually all prior related work has used data on publicly traded companies or non-representative samples of private companies, even as private companies comprise the vast majority of U.S. corporations, account for approximately half of C corporation employment, and contribute significantly to innovative activity in the economy. Second, our analysisofthecomponentsofthecorporatetaxreformprovidesvaluableinformationonthescope and relative economic significance of different provisions of TCJA. Doing so sheds new light on the breadth and depth of reforms that comprise TCJA, beyond the most conspicuous corporate incometaxratechanges. Thispaperalsocontributestothebroadbodyofresearchineconomics,finance,andaccounting that studies trends in and determinants of corporate tax rates and tax-related financial positions 2

both domestically and globally (Slemrod, 2004; Clausing, 2007; Dyreng et al., 2017; Gaertner, Glover, and Levine, 2021; Wier and Zucman, 2022). In related studies examining corporate tax trends around TCJA, Dowd, Giosa, and Willingham (2020) use IRS data and document that large firmsengagedinconsiderablebehavioralandtax-shiftingresponsestothelegislation. Dyrengetal. (Forthcoming)andWagner,Zeckhauser,andZiegler(2020)usedatafrompublicfinancialfilingsto studytrends inglobalETRs aroundTCJAandexamine theassociationbetween ETRchangesand firmcharacteristics. HenryandSansing(2018)studyscaledcashtaxdifferencesforpublicfirmsin additiontoETRsandfindthatTCJAhadnoeffectontax-favoredstatusforpublicfirmsoverallbut reduced the share of tax-favored profitable firms.2 Several concurrent papers also examine certain provisions of TCJA in isolation, including the interest limitation (Carrizosa, Gaertner, and Lynch, 2020) and the international provisions (Clausing, 2020; Garcia-Bernardo, Jansky`, and Zucman, 2022;Albertus,Glover,andLevine,2023). Our work is also related to the growing literature studying effects of TCJA on firm behavior.3 The majority of the current research studies effects of TCJA on public firm behavior. However, research on private firms suggests considerable differences in tax-related incentives and behavior ofprivatefirmscomparedtopublicfirms(MillsandNewberry,2001;Grahametal.,2014;Hoopes et al., 2020; Dobridge, Lester, and Whitten, 2021). Given the different post-TCJA trends that wedocumentintaxpositionsforprivatedomesticfirms,ourworkraisesthequestionofhowother outcomesmayhavedifferedforprivatefirmscomparedtopublicfirmsafterTCJAaswell. Overall, our results underscore that TCJA’s policy effects for private domestic firms are likely to be very differentfrompublicfirmsandmultinationals. I Overview of TCJA Corporate Tax Provisions In this section, we provide a short summary of the numerous domestic and international corporate taxchangesinTCJA.4 Beginningwiththedomesticbusiness-relatedchanges,theJointCommittee 2Thescaledcashtaxdifferenceisdefinedasyear-to-yearchangesincashtaxespaiddividedbythemarketvalue ofassets,whichcanbecalculatedforfirmswithlosses. Theauthorsdefinetax-favoredstatusasifafirm’scashETR continuedtobelessthanthestatutoryrateorifscaledcashtaxdifferencescontinuedtobenegativeafterTCJA. 3See, for example, work on corporate actions and statements following TCJA enactment (Hanlon, Hoopes, and Slemrod, 2019), contributions to defined benefit pension plans (Gaertner, Lynch, and Vernon, 2020), changes in executive compensation (De Simone, McClure, and Stomberg, 2022), reclassification of business costs (Laplante etal.,2021),organizationalstructure(Henry,Plesko,andUtke,2018),earningsmanagement(Kubick,Lockhart,and Robinson,2021),bankdepositratesandlending(FoxandPyle,2022),andemployment,workerearnings,andcapital investment (Kennedy et al., 2022). Auerbach (2018) discusses predicted economic impacts of TCJA and Gale et al. (2019) examines TCJA’s effects on aggregate domestic activity, including nonresidential fixed investment, non-farm employment,andrealearnings. 4A detailed description of all the statutory changes enacted under TCJA and an explanation of the prior law’s provisions are presented in Joint Committee on Taxation (2018), estimated budget effects are included in Joint CommitteeonTaxation(2017),additionaldiscussionofthelegislativemotiveforthechangesiscontainedinDowd, Giosa,andWillingham(2020),anddiscussionofaspectsofthepolicydebateisincludedinAuerbach(2018). 3

on Taxation estimated these provisions would reduce federal revenue by $650 billion over the 10-yearbudgetwindowfollowingTCJA—fiscalyears2018to2027. Themostsubstantialchange, in terms of revenue consequences, was the change in the corporate tax rate schedule from an approximately graduated statutory tax schedule, with a top marginal tax rate of 35% on income in excess of $18 million, to a flat tax schedule with a rate of 21%. TCJA also repealed the corporate alternative minimum tax (AMT) while allowing firms to offset tax liability or receive a credit for unusedAMTcredits. ThreeaspectsofTCJAdirectlyaffecteddeductionsforcapitalexpenditures,interestexpenses, and business activity associated with domestic production. First, under TCJA, the depreciation schedule for most capital investment was changed to full expensing through 2022. Second, TCJA imposed a new limit on the deductibility of interest expenses to 30% of adjusted taxable income.5 Third, TCJA repealed the domestic production activity deduction (DPAD), which had provided taxpayers a deduction of 9% for income from qualifying activities related to producing goods and servicesdomestically.6 Therewerealsokeyprovisionsrelatedtothetaxtreatmentoflosses. PriortoTCJA,firmswere permitted to apply NOLs against taxable income for 2 years prior (“carryback”) or apply NOLs against taxable income as many as 20 years into the future (“carryforward”). TCJA eliminated NOL carrybacks and changed the limitation on carryforwards to 80% of pre-NOL taxable income for20years. TCJA made extensive changes to tax provisions related to foreign income and operations as well. Toprovideasenseofscale,theJCTestimatedthatthetotalinternationaltaxreformchanges wouldraiseabout$325billionfrom2018to2027. First, the legislation changed the treatment of foreign income from a worldwide system of taxation, whereby foreign earnings were generally taxed only when they were repatriated to the United States, to a territorial system that eliminates the tax on repatriated or unrepatriated foreign earnings (i.e., gives a 100% deduction for dividends received from foreign subsidiaries). The legislation also included a one-time transition tax on the previously untaxed earnings of U.S. multinationalsof8%onnon-cashassetsand15.5%oncashassetsheldoversees. Second,severalnewinternationaltaxprovisionswereimplementedrelatedtofirmincome-shifting incentives, particularly: the base erosion and anti-abuse tax (BEAT), the global intangible low-taxed income (GILTI) provision, and the foreign derived intangible income (FDII) provision. Theswitchtoaterritorialinternationaltaxsystemincreasedincentives,tosomeextent,forfirmsto moveoperationsandincomeoverseasbecauseanytaxsavingsfromdoingsowouldbepermanent 5Through 2021, adjusted taxable income was defined as taxable income excluding business interest income and expense,depreciation,amortization,depletion,andNOLs.After2021,adjustedtaxableincomewasdefinedastaxable income excluding business interest income and expense and NOLs. Other TCJA provisions related to deductions included the limitation of deductions for fringe benefits and expanding the definition of executive compensation for thepurposesofSection162(m). 6SeeDobridge,Landefeld,andMortenson(2021)andOhrn(2018)fordetaileddiscussionsoftheDPAD. 4

instead of deferred. The BEAT, GILTI, and FDII were intended to counteract these effects and decreaseincentivestomovebusinessactivityoverseas. TheBEATimposedaminimum10percent U.S. tax on modified taxable income, which excluded deductible transactions made between relatedparties(i.e.,paymentsmadefromaU.S.parenttoacontrolledforeigncorporation). GILTI wasintendedtoreduceincentivesforfirmstorelocateoperationstolower-taxcountriesbylevying aminimumtaxonafirm’sforeignearningsgreaterthan10%oftotalforeigntangibleassets. FDII reduced firm taxes on U.S. earnings derived from foreign sales and was intended to incentivize firmstolocateintangiblecapitaldomestically. II Description of the Corporate Tax Data The corporate tax return data we use for this analysis are sourced from Form 1120 corporate tax filings and related schedules, as provided by the Statistics of Income (SOI) Division of the Internal Revenue Service (IRS). In particular, we use data from a stratified random sample of C corporation tax returns that is created, cleaned, and edited each year by the SOI (U.S. Internal Revenue Service, 2013). We refer to these data as the “SOI sample.” We focus on tax years 2010 to 2019, corresponding to the period after the Global Financial Crisis in 2007-2009 but before the Covid-19pandemicin2020. TheSOIsamplefrom2010-2019includes,onaverage,approximately 54,000 firms per tax year. For the main analysis, we use data on tax payments from the front page ofForm1120anddataonbookincomefromSchedulesM-1andM-3.7 Our primary variable of interest is the domestic U.S. corporate effective cash tax rate (ETR), which we define as total taxes paid over domestic pre-tax book income. We calculate this for all firms with positive book income, following the approaches of Hoopes et al. (2020) and Dobridge, Lester,andWhitten(2021).8 Detaileddefinitionsforallvariablesusedinthisstudy,includingIRS formandlinenumbers,arereportedinAppendixA. WesplittheSOIsamplealongseveralfirmcharacteristicstostudytheheterogeneouseffectsof TCJA.Thesecharacteristicsincludepublicownership,privateownership,multinationaloperations, domestic-only operations, and size bins of domestic sales. We define a firm as public if it ever 7We exclude from the SOI corporate sample filings of Forms 1120-S, 1120-L, 1120-RIC, 1120-F, 1120-REIT, and 1120-PC. Note that total taxes paid will include some U.S. taxes on foreign source income which is subject to tax. PriortoTCJA,thisincludeddividendsfromcontrolledforeigncorporations(CFCs),foreignbranchincome,and subpartFincome. AfterTCJA,thiscouldalsoincludetaxowedonGILTIandtaxontheone-timetransitiontaxfor unrepatriatedforeignearnings. 8Specifically, when a firm reports attaching a Schedule M-3 to the Form 1120 (Box A4), we calculate the ETR as taxes paid (Form 1120, line 31) divided by pre-tax financial statement income, where pre-tax financial statement incomeisthesumofnetincome(ScheduleM-3,PartI,line11),U.S.currentincometaxexpense(ScheduleM-3,Part III, line 1), and U.S. deferred income tax expense (Schedule M-3, Part III, line 2). However, only firms with assets above$10millionarerequiredtofiletheScheduleM-3. WhenafirmdoesnotreportattachingaScheduleM-3,we definepre-taxfinancialstatementincomeasthesumofnetincomeperbook(ScheduleM-1,line1)andfederalincome taxperbooks(ScheduleM-1,line2). 5

reports having publicly traded voting common stock on Schedule M-3, line 3a, during the sample period from 2010 to 2019, and categorize all other firms as private firms.9 We define a firm as multinational if it ever reports having a controlled foreign corporation (CFC) (i.e., a foreign subsidiary)byfilingForm5471duringthesampleperiod. Finally,tostudyfirmsofdifferentsizes, we separate them across categories of the firm domestic sales distribution, using total domestic grossreceiptsreportedonline1cofForm1120. We construct several different analysis samples of firms in this paper. For our analysis of the time trends in U.S. domestic cash ETRs, we require a firm to have non-negative and non-zero book income in a given year to be included in the sample for that year, such that we are able to calculate a cash ETR for each firm (designated “ETR sample”). For our analysis of changes in the domestic cash ETR from 2016 to 2019, we require a firm to have a non-missing cash ETR in both of those years (designated “ETR change sample”). Finally, when we examine firms’ likely exposure to various TCJA provisions based on pre-TCJA characteristics, we study the full SOI sampleoffirms. Panel A of Table 1 presents summary statistics for the ETR sample of firm characteristics in 2016,priortoTCJAenactment,formultinational,domestic,public,andprivatefirms. Allsamples are weighted using SOI sampling weights to produce firm-weighted population averages. The panel highlights stark average differences across firm types in variables such as receipts, assets, taxes paid, and employment, with public multinationals being the largest and private domestic firms being the smallest. Online Appendix Figure B1 shows the industry distribution of firms in the sample by firm type. Panel A of Table 1 also documents the mean ETRs for 2016 and 2019 by firm-type.10 The decrease in mean ETRs for public firms and private multinationals is in stark contrasttotheincreaseforprivatedomesticfirms. PanelBemphasizesthesedifferencesfurther,reportingtheshareofeachfirmtypeinaggregate firm counts, sales, taxes, and employment. Among C corporations, public multinationals account for a small share of firms (less than 1%) but account for approximately half of gross receipts, two-thirdsoftaxespaid,and45%ofemployment. Bycontrast,privatedomesticbusinessesaccount forover98%offirmsbutaccountforapproximately22%ofgrossreceipts,14%oftaxespaid,and 31%ofemployment. 9This method of identifying publicly traded firms misclassifies as private any publicly traded firm that was not requiredtofileascheduleM-3. 10DuetolargeoutliersinETRs,theETRandchangeinETRarewinsorizedatthe5%levelthroughoutthepaper. 6

Table1: SummaryStatistics PanelA:2016MeansbyFirmType (1) (2) (3) (4) Public Public Private Private Multinational Domestic Multinational Domestic Sales(000s) 5,135,210 451,470 361,645 4,711 Totalassets(000s) 22,839,687 6,105,553 677,701 7,141 Depreciableassets(000s) 3,132,429 477,687 161,130 1,752 Payroll(000s) 843,377 86,666 53,704 888 Netincome(000s) 529,296 42,473 20,470 224 Taxespaid(000s) 116,925 12,547 5,048 61 U.S.employment 14,421 2,151 1,027 26 ETRin2016 17.5 17.5 18.8 12.0 ETRin2019 10.7 11.9 12.6 15.1 Uniquefirms 1,542 1,345 8,081 602,717 PanelB:2016ShareofAggregatesbyFirmType (1) (2) (3) (4) Shareof Public Public Private Private CCorp Multinational Domestic Multinational Domestic #Firms 0.002 0.001 0.012 0.985 GrossReceipts 0.517 0.039 0.222 0.221 TaxPaid 0.665 0.051 0.147 0.136 Employment 0.454 0.057 0.176 0.314 PanelAreports2016samplemeansoffirmcharacteristicsforfirmsintheETRsample,separatelybyfirmtype. Panel Breportstheshareofeachfirmtypein2016aggregatefirmcounts, sales, taxespaid, andemployment. Thedataset wascreatedbytheauthorsusingIRSadministrativetaxdata,andvariablesaredefinedinAppendixA.Thedataare weightedusingSOIsamplingweightstoreflecttheU.S.populationofCcorporations. 7

III Results III.A Trends in Domestic Cash ETRs WebeginouranalysisbypresentingtrendsinaverageETRsfrom2010to2019inFigure1. Figure 1a separately displays ETRs for different types of firms: public multinational, public domestic, privatemultinational,andprivatedomestic. Figure1bpresentsETRsatdifferentpointsinthesize distribution by total sales: firms above the 95th percentile, firms in the 90th to 95th percentile, and firms below the 90th percentile. As in Table 1, all samples are weighted using SOI sampling weightstoproducefirm-weightedpopulationaverages. In Figure 1a, we observe a similar trend in average firm domestic cash ETRs for public C corporations as documented by Dyreng et al. (Forthcoming) for the average firm global cash ETR of public U.S. firms. After fluctuating between about 17% and 18% between 2012 and 2017, cash ETRsforpublicmultinationalanddomesticfirmsdeclinedsharplyin2018afterpassageofTCJA, to about 11%. In 2019, ETRs declined further for public multinationals (to around 10%) and increased somewhat for public domestic C corporations (to around 13%). Private multinational firms experienced a similar change as public multinational firms, though their average ETRs were oneortwopercentagepointshigherthanpublicfirmsinthefiveyearsleadinguptoTCJApassage, and they experienced a somewhat smaller drop in ETRs after passage, to about 13% in 2018 and 12%in2019.11 Trends in the ETRs of private domestic firms were starkly different than those of public firms or private multinationals. After a gradual increase in ETRs over most of the pre-TCJA sample period, ETRs increased discontinuously in 2018 and ticked up again in 2019. By 2019, the averageETRfordomesticprivatefirmswashigherthanforpublicfirmsorforprivatemultinational corporations—about 14% for private domestics compared to about 10% for public multinationals, for example—a striking reversal of the pre-TCJA trend when average ETRs were considerably lowerforprivatedomesticfirmsthanforpublicfirmsandprivatemultinationals.12 Figure 1b decomposes trends by firm size, and illustrates that the largest tax cuts were concentrated amongst the largest firms. For firms in the top five percent of the sales distribution, averageETRsdeclinedfromabout23%in2016to16%in2019. Firmsinthe90to95thpercentile 11Forcompleteness,inOnlineAppendixFigureB2wealsoreporttimeseriesextendingtotaxyear2020;thedata from2020shouldbeinterpretedwithcautionandcareduetotheeconomiceffectsandpolicyresponsesrelatedtothe pandemic-inducedrecession. 12AscashETRscanonlybecalculatedforobservationswithpositivebookincome,theOnlineAppendixcontains figuresdisplayingtrendsinacomplementarymeasureoffirmtaxpositionsthatcanbecalculatedforthefullSOIfirm sample: thefractionoffirmsthatpaytaxes,byfirmtype(FigureB3a)andfirmsize(FigureB3b). Consistentwiththe ETRanalysis,weobservethatthefractionofpublicfirms,privatemultinationalfirms,andlargefirmsthatpaytaxes declinesnotablyafterpassageofTCJA.Incontrast,thefractionofprivatedomesticfirmsandsmallfirms(thebottom 90th oftheemploymentsizedistribution)thatpaytaxesincreasedabitfrom2017to2018androsesomewhatfurther in2019. 8

in terms of sales had relatively small average ETR declines, from 17% in 2016 to 16% in 2019, whilesmallfirmsexperiencedanaverageETRincrease. 9

Figure1: EffectiveTaxRatesforCCorporations: 2010to2019 Overall Public MNE Public Domestic Private MNE Private Domestic Percent 20 18 16 14 12 10 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 (a)ByPublicandMultinationalStatus Overall Bottom 90% 90-95% 95%+ Percent 25 20 15 10 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 (b)BytheDistributionofSales This figure presents trends in average domestic cash effective tax rates from 2010 to 2019 for U.S. C corporations. Figure 1a presents trends for firms overall, for public multinationals (MNEs), public domestic firms, private MNEs andprivatedomesticfirms. Figure1bshowstrendsforfirmsoverall,forfirmsinthe95thpercentileofthefirmsales distribution in 2016, firms in the 90th to 95th percentile of the sales distribution in 2016, and firms below the 90th percentilein2016. AllpanelswerecreatedbytheauthorsusingIRSadministrativetaxdataandvariablesaredefined inAppendixA.Bubblesizesareproportionaltotheaggregateshareof2016federalcorporatetaxrevenuecollected fromeachfirmtype. AllsamplesareweightedusingSOIweightstoproducefirm-weightedpopulationaverages. 10

Figure2: DistributionofETRs: 2016and2019 40 30 20 10 0 smriF fo tnecreP 2016 2019 0 10 20 30 40 50 Percent (a)FullSample 30 20 10 0 smriF fo tnecreP 2016 2019 25 20 15 10 5 0 0 10 20 30 40 50 Percent (b)PublicMultinational smriF fo tnecreP 2016 2019 0 10 20 30 40 50 Percent (c)PublicDomestic 40 30 20 10 0 smriF fo tnecreP 2016 2019 40 30 20 10 0 0 10 20 30 40 50 Percent (d)PrivateMultinational smriF fo tnecreP 2016 2019 0 10 20 30 40 50 Percent (e)PrivateDomestic ThisfigurepresentsthedistributionofETRspriortoTCJApassage,in2016,andfollowingpassage,in2019. Figure 2ashowsthedistributionforallfirms,andFigures2b,2c,2d,and2eshowthedistributionsforpublicmultinationals, publicdomesticfirms,privatemultinationals,andprivatedomesticfirms,respectively. Allpanelswerecreatedbythe authors using IRS administrative tax data and variables are defined in Appendix A. All samples are weighted using SOIweightstoproducefirm-weightedpopulationaverages. 11

To further investigate dispersion in ETR changes across firms, we examine the distribution of ETRs before and after TCJA. Figure 2a presents a histogram of the distribution of ETRs in 2016 and 2019. We focus on 2016 and 2019 as our points of comparison due to evidence that firms engagedintax-shiftingbehaviorsin2017and2018tominimizetheirtaxliabilities(Dowd,Giosa, and Willingham, 2020). We find that annual ETR averages mask considerable variation in ETRs across firms. In both 2016 and 2019, the distribution of ETRs is bimodal, with the most frequent observationsaroundzeroETRinbothyears,reflectingthelargefractionoffirmswithpositivebook incomebutzeroU.S.federalcashtaxespaid. In2016,thedistributionshowsasecondmassaround 15%, corresponding to the marginal tax rate for firms with taxable income less than $50,000.13 In 2019, the distribution shows a second mass around 21%, corresponding to the newly introduced, flatcorporatetaxrateforallCcorporations. Figures2bthrough2eshowthedistributionofETRsin2016and2019byfirmtype. Consistent with Figure 1, the distribution of ETRs shifted to the left for public firms and for private multinationals, but shifted to the right for private domestic firms. Given that private domestic firms make up the majority of U.S. C corporations, the figure documents an important fact about the distribution of ETRs before and after TCJA: the vast majority of U.S. C corporations with positive book income experienced an ETR increase between 2016 and 2019 (approximately 81% of all C corporations), despite the reduction in marginal tax rates for most of the pre-TCJA tax brackets.14 The picture is considerably different for public firms and private multinationals, where the distribution of ETRs shifts to the left in 2019. This picture is similar to findings for public multinationals documented by Wagner, Zeckhauser, and Ziegler (2020) for the distribution of changes in firms’ global cash ETRs after TCJA. In addition, the mass of public and private multinationalsatazeroETRincreasedin2019relativeto2016. III.B Exposure to TCJA Corporate Provisions In this section, we move beyond studying the ETR—an “overall” measure of tax liability—and examine the scope and economic salience of specific TCJA provisions. Given the lack of transparency and scarcity of available granular information about many firms’ financial and tax positions, it can be challenging for researchers and policymakers to understand the exposure of differenttypesoffirmstothevariouslegislativechangesinthebill. Tofillthisgapintheliterature, we document firms’ tax-related characteristics in 2016, one year prior to the passage of TCJA and two years prior to most provisions taking effect, as a proxy for exposure to different provisions 13Averageandmarginaltaxratesarethesameinthefirstbracketofagraduatedratestructure. 14Weapproximatetheshareoffirmsreceivingataxincreasebymatchingfirmswithpositivebookincomefromthe 2016 and2019 SOI samplebased on theirsampling weight, whichis a functionof the firm’sassets, and comparing thedifferenceintheirETRs. Wethencomputetheweightedshareoffirmsthatreceivedataxincrease. 12

priortoTCJA. Table 2 reports sample means for a list of firm-level indicators that serve as proxy measures for exposure to TCJA’s major provisions. The fraction of exposed firms is presented for all firms in column 1, public multinationals in column 2, public domestic firms in column 3, private multinationalsincolumn4,andprivatedomesticfirmsincolumn5. ThecenterpieceofTCJAcorporateprovisionswasthechangefromanapproximatelygraduated corporate tax schedule, with a maximum rate of 35%, to a flat corporate income tax rate of 21%. Tobetterunderstandtherelativeimportanceofthecorporatetaxratechange,weexaminetheshare of firms in each tax bracket in 2016. From that baseline, a substantial fraction of private domestic U.S. firms would have had higher ETRs under TCJA’s new tax rate schedule (ceteris paribus): 43% were in the first tax bracket (with a 15% tax rate) in 2016 and 36% of private domestic firms paid no tax. In contrast, virtually 0% of public firms were in the first tax bracket while a similar fraction of private multinationals and public firms paid no tax. For firms with positive taxable income,thelargestfractionofpublicdomesticandprivatemultinationalswereinthe34%bracket (with between $335 thousand and $10 million in taxable income) and for public firms, 57% were inthetop35%taxbracket,withgreaterthan$18millionintaxableincome. The corporate AMT repeal and DPAD repeal were more likely to affect multinationals and public firms than domestic private firms. In 2016, the AMT affected few private domestic firms (about 1%), but around 12% and 9% of public and private multinationals, respectively, and about 16% of public domestic firms. The DPAD was most utilized by multinationals, claimed by 45% and 24% of public and private multinationals in 2016, respectively, compared to just 9% of public domesticfirmsand6%ofprivatedomesticfirms. The other two provisions with wide applicability were the NOL limitations and the capital expensing provisions. Sixty-one percent of public multinationals and 50% of public domestic firms were able to claim an NOL carryforward, as well as 43% and 46% of multinational and domestic private firms, respectively. A large majority of public firms and private multinational firmsalsoclaimeddepreciablecapitalexpendituresin2016,whichwouldlikelyhavebeeneligible for immediate, full bonus depreciation (and lower current-year taxable income) under TCJA. This was more common for public multinationals and domestic firms (94% and 85%, respectively) than private multinationals (73%). Of private domestic firms, 38% claimed depreciable capital expendituresin2016.15 Theinterestlimitation—whichwasbasedontheamountofinterestexpenserelativetoadjusted taxable income—would have affected a small share of U.S. firms. While a large fraction of firms had positive interest expenses, a small fraction had positive taxable income and interest expense abovethelimitationin2016: around5%ofpublicandprivatemultinationals,forexample. 15NotethatpriortoTCJA,manysmallerfirmswerealreadyabletoexpenseaportionoftheircapitalexpenditures throughSection179. 13

Table2: ExposuretoTCJAProvisions (1) (2) (3) (4) (5) TCJA All Public Public Private Private Provision Firms Multinational Domestic Multinational Domestic IncomeTaxBracket Zerotaxableincome,>0netincome 0.310 0.173 0.210 0.251 0.311 Zerotaxableincome,≤0netincome 0.050 0.120 0.135 0.077 0.049 15%taxbracket(<50Ktaxableincome) 0.427 0.000 0.001 0.043 0.435 25%bracket($50K-75Ktaxableincome) 0.053 0.001 0.001 0.006 0.054 34%bracket($75K-100Ktaxableincome) 0.032 0.000 0.008 0.012 0.033 39%bracket($100K-335Ktaxableincome) 0.069 0.005 0.029 0.119 0.069 34%bracket($335K-10Mtaxableincome) 0.052 0.086 0.282 0.340 0.047 35%bracket($10M-15Mtaxableincome) 0.001 0.026 0.053 0.033 0.001 38%bracket($15M-18Mtaxableincome) 0.001 0.016 0.027 0.013 0.000 35%bracket(>$18Mtaxableincome) 0.005 0.573 0.255 0.106 0.001 ExposuretoSelectDeductions PaidthecorporateAMT 0.015 0.124 0.162 0.092 0.013 HadanNOLcarryforward 0.456 0.608 0.501 0.433 0.455 UsedanNOLcarryforward 0.430 0.513 0.385 0.382 0.431 Depreciablecapitalexpenditures 0.388 0.936 0.845 0.725 0.381 ClaimedtheDPAD 0.064 0.456 0.087 0.244 0.060 Aboveinterestlimit,>0taxableincome 0.031 0.057 0.014 0.045 0.031 Positivenetinterestexpense 0.420 0.764 0.325 0.636 0.416 ExposuretoInternationalProvisions Unrepatriatedforeignearnings 0.005 0.593 0.000 0.290 0.000 PositiveforeignE&P 0.010 0.834 0.000 0.587 0.000 ReceiveddividendsfromCFCs 0.003 0.298 0.033 0.094 0.001 ForeignE&PinexcessofDTIR 0.006 0.547 0.000 0.327 0.000 ThetablepresentsdescriptivestatisticsoftheshareoffirmsexposedtovariousTCJAcorporatetaxprovisionsin2016, priortothelaw’senactment. Statisticsareprovidedforallfirmsincolumn(1)andfourgroupsoffirmsincolumns(2) to(5): Publicmultinationals,publicdomesticfirms,privatemultinationals,andprivatedomesticfirms. Statisticsare firm-weightedusingSOIweightstoberepresentativeoftheU.S.populationofCcorporations. Thetablewascreated bytheauthorsusingIRSadministrativetaxdataandvariablesaredefinedinAppendixA.Allsamplesareweighted usingSOIweightstoproducefirm-weightedpopulationaverages. 14

Lastly, we examine the relative importance of the corporate international provisions.16 A cornerstone of the multinational policy changes was the switch from a worldwide tax system, whereby foreign earnings were taxed when repatriated to the United States, to a territorial tax system, whereby foreign earnings would be taxed at foreign, local rates. Approximately 83% of public multinationals and 59% of private multinationals had positive foreign earnings overall in 2016. About 30% of public multinationals and 9% of private ones repatriated earnings to the United States in that year, which manifests as a U.S. parent receiving a dividend from a CFC. Finally, we also evaluate the global intangible low-taxed income (GILTI) provision. GILTI exacts atleasta10.5%taxonafirm’sforeignearningsthataregreaterthan10%ofqualifiedbusinessasset investments (QBAI). In 2016, we observe that 55% of public multinationals and 33% of private multinationalshadforeignearningsinexcessof10%oftangibleassets,aproxyforQBAI. III.C Correlates of ETR Changes How was exposure to various TCJA provisions correlated with changes in firms’ effective tax rates? This section estimates the relative salience of different provisions to observed ETR changes documented in Section III.A. We regress firm-level ETR changes from 2016 to 2019 on indicatorsforpre-TCJAfirm-levelexposuretosignificantprovisionsofTCJA,usingthefollowing specification: ∆𝐸𝑇𝑅 = 𝛼+𝛽 · ≤ 2𝑛𝑑𝑀𝑇𝑅𝐵𝑟𝑎𝑐𝑘𝑒𝑡 +𝛽 · ≥ 4𝑡ℎ𝑀𝑇𝑅𝐵𝑟𝑎𝑐𝑘𝑒𝑡 +𝛽 𝐴𝑀𝑇 +𝛽 𝑁𝑂𝐿 𝑓 1 𝑓 2 𝑓 3 𝑓 4 𝑓 +𝛽 𝐵𝑜𝑛𝑢𝑠 +𝛽 𝐷𝑃𝐴𝐷 +𝛽 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝐿𝑖𝑚𝑖𝑡𝑎𝑡𝑖𝑜𝑛 +𝛽 𝑀𝑁𝐶 +𝜖 (1) 5 𝑓 6 𝑓 7 𝑓 8 𝑓 𝑓 Intheregression,theoutcome∆𝐸𝑇𝑅 isthe2016to2019changeinthecashETR,measured 𝑓 at the firm level (𝑓), and all the regressors are defined using firm characteristics in 2016. ≤ 2𝑛𝑑𝑀𝑇𝑅𝐵𝑟𝑎𝑐𝑘𝑒𝑡 is an indicator variable equal to 1 if a firm was in the first (15%) or second (25%) tax bracket or paid zero tax as defined in Table 2; ≥ 4𝑡ℎ𝑀𝑇𝑅𝐵𝑟𝑎𝑐𝑘𝑒𝑡 is an indicator variable equal to 1 if a firm was in the fourth tax tax bracket or higher as defined in Table 2; 𝐴𝑀𝑇 is an indicator variable equal to 1 if a firm paid the corporate AMT, 𝑁𝑂𝐿 is an indicator variable equal to 1 if a firm used an NOL carryforward; 𝐵𝑜𝑛𝑢𝑠 is an indicator variable equal to 1 if a firm reported any capital expenditures on Form 4562; 𝐷𝑃𝐴𝐷 is an indicator equal to 1 if a firm claimed the DPAD; 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝐿𝑖𝑚𝑖𝑡𝑎𝑡𝑖𝑜𝑛 is a dummy variable equal to 1 if a firm is above the interest limitation and had positive taxable income (denoted 𝐴𝑏𝑜𝑣𝑒 𝑖𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝑙𝑖𝑚𝑖𝑡𝑎𝑡𝑖𝑜𝑛, > 𝑡𝑎𝑥𝑎𝑏𝑙𝑒 𝑖𝑛𝑐𝑜𝑚𝑒inTable2);and𝑀𝑁𝐶 isadummyvariableequalto1ifafirmhasaCFCduring thesampleperiod. TheregressionisweightedusingSOIsamplingweightsandthestandarderrors 16The international provisions generally may lead to a change in the amount of U.S. corporate tax owed by a corporationwithoutanychangeintheirdomesticbookincome. 15

areclusteredbyfirm. Thecoefficientestimatesfromthisspecificationcapturenotonlycorrelationsbetweenexposure to TCJA’s provisions and changes in ETRs, but may also reflect firms’ behavioral responses to the policies that generate secondary effects on their ETRs. For example, while the direct effect of a cut in the corporate income tax rate may be to reduce the firm’s ETR, the firm in response may engage in behaviors that increase its taxable profits, which in turn may have the second-order effect of increasing the firm’s tax liability and offsetting the ETR cut. We abstract from these second-order effects in the following analysis but note that firms’ behavioral responses to similar policyprovisionshavebeenstudiedinKennedyetal.(2022),Dobridge,Landefeld,andMortenson (2021),Dowd,Giosa,andWillingham(2020),andZwickandMahon(2017),amongothers. Figure3: CoefficientEstimates: TCJAprovisionsandETRChanges Outcome: 2016-2019 Change in Firm ETR (pp) 2016 Firm Characteristic <=2nd MTR Bracket >=4th MTR Bracket AMT NOL Bonus (Inv>0) DPAD Interest Limitation Multinational -10 -5 0 5 N = 15,889 SOI firms Constant: -2.43, se=1.00 Thefigurepresentscoefficientestimatesand95%confidenceintervalsofaregressionofthechangeinETRsfrom2016 to 2019 on indicators measuring a firm’s exposure in 2016 to various future TCJA provisions, as defined in Section III.C. ThefigurewascreatedbytheauthorsusingIRSadministrativetaxdata. TheregressionisweightedusingSOI sampling weightsand is tabulated in OnlineAppendix Table B1. Furtherinformation aboutvariable construction is includedinAppendixA. Figure3displaystheestimatedcoefficientsand95%confidenceintervalsforthepolicy-specific indicator variables in equation 1.17 These coefficient estimates suggest exposure to the statutory marginaltaxratechangesandtheNOLchangeshavethelargestconditionalcorrelationswithETR 17The regression results are tabulated in Online Appendix Table B1. The table also presents results including two-digitindustryfixedeffectsinthespecification,whichareverysimilar. 16

changes. Being in the first or second MTR bracket or paying no tax in 2016 is associated with about a four percentage point larger increase in ETR from 2016 to 2019, compared to firms in the third MTR bracket (the 34% bracket—the omitted group in the regression).18 By contrast, being in the fourth MTR bracket or above in 2016 is associated with about a 91 percentage point ETR cut 4 compared to firms in the third bracket. Using an NOL carryforward in 2016 is associated with about a 41 percentage point higher ETR change. The coefficient on pre-TCJA AMT exposure has 2 the expected negative association with the ETR change, although the coefficient on the interest limitation is not positive as expected; in both cases, however, the magnitudes are smaller than the income tax rate changes and NOL changes, and the correlations are not statistically significant. Exposure to bonus depreciation had almost no conditional correlation with ETR changes, while DPAD exposure was weakly and positively correlated. Multinational firm status was associated with a 2 percentage point reduction in ETR.19 Overall, the results provide suggestive evidence that TCJA’s changes to the corporate tax rate schedule and to the tax treatment of NOL’s were particularly economically relevant for explaining changes in firms’ tax positions in the years immediatelyfollowingthepolicychange. IV Conclusions In this paper, we use a representative sample of U.S. C corporation tax returns to summarize the direct effects of TCJA corporate tax provisions on the tax positions of corporations. While the declineinpublicfirmETRsafterTCJAhasbeenwell-documentedbyotherresearchersandTCJA’s corporate tax cuts are commonly discussed in media and other public venues (e.g., see Rubin and Francis (2018) and Rubin and Francis (2021)), we present several novel facts about changes in ETRsandtherelativeimportanceofseveralaspectsofthiscomplexreform. We document that a majority of U.S. firms with positive book income saw an increase in their ETR after passage of TCJA and that the average domestic ETR increased for U.S firms overall as well. Both trends were driven by an increase in ETRs for private domestic firms. We show that similar to public MNCs, private MNCs saw a decline in ETRs after TCJA, but a smaller decline thanwasexperiencedbypublicmultinationalanddomesticfirms. Asof2019,theaverageETRfor private domestic firms with positive book income was higher than for public firms or for private 18Thethirdbracketwasomittedbecauseitisthefirstwithanunambiguousreductionintheaverageandmarginal statutory tax rates. The point estimate on the constant term indicates firms in that bracket had an ETR reduction of around21 percentagepoints. 2 19Thecomplexnatureoftheinternationalreformsmakesitdifficulttoformanex-antepredictiononthesignofthe multinational status indicator. Some provisions, like FDII, were unambiguously rate cuts and some, like the BEAT, wereunambiguouslyincreases. GILTI,ontheotherhand,mayhaveactedasaratecutforsomeandanincreasefor others. TheoriginalJCTscoreoftheinternationalprovisionsofTCJAindicatedaslightrevenuelossexcludingthe repatriationtaxundersection965. 17

multinationals—a striking reversal of the pre-TCJA ETR positions of these firms. Finally, we find that the change in statutory tax rates and the limitation on net operating losses had the strongest correlationwithchangesinfirmETRsfrom2016to2019. ExposuretothecorporateAMTrepeal, the interest limitation, and the multinational provisions had smaller, negative correlations, while exposuretothebonusdepreciationandDPADchangeshadlittleeffect. Our work contributes to the body of existing research by presenting a more comprehensive picture than previous work of the effects of TCJA on tax positions of U.S. corporations, the majority of which are privately held, domestically based, and small in terms of total assets and employment. Given the opaqueness of tax positions of privately held firms that do not file public financial statements, as well as limited information even about tax positions included in public firm financial statements, it is challenging to evaluate changes in the U.S. tax code. This study provides a resource for researchers, policymakers, and the general public to better understand the directeffectsofthissweepingreform,includingitscomponentparts. 18

References Albertus, James F, Brent Glover, and Oliver Levine. 2023. “The Real and Financial Effects of InternalLiquidity: EvidenceFromtheTaxCutsandJobsAct.” AvailableatSSRN4471259. Auerbach, Alan J. 2018. “Measuring the effects of corporate tax cuts.” Journal of Economic Perspectives32(4):97–120. Carrizosa,RichardD.,FabioB.Gaertner,andDanielP.Lynch.2020. “Debtandtaxes? Theeffect ofTCJAinterestlimitationsoncapitalstructure.” JournaloftheAmericanTaxationAssociation 45(2):1–22. Clausing,KimberlyA.2007. “CorporatetaxrevenuesinOECDcountries.” InternationalTaxand PublicFinance14(2):115–133. ———. 2020. “Profit shifting before and after the Tax Cuts and Jobs Act.” National Tax Journal 73(4):1233–1266. De Simone, Lisa, Charles McClure, and Bridget Stomberg. 2022. “Examining the effects of the Tax Cuts and Jobs Act on executive compensation.” Contemporary Accounting Research 39(4):2376–2408. Decker, Ryan, John Haltiwanger, Ron Jarmin, and Javier Miranda. 2014. “The role of entrepreneurship in US job creation and economic dynamism.” Journal of Economic Perspectives28(3):3–24. Dobridge, Christine L., Paul Landefeld, and Jacob Mortenson. 2021. “Corporate taxes and the earningsdistribution: EffectsoftheDomesticProductionActivitiesDeduction.” WorkingPaper . Dobridge,ChristineL.,RebeccaLester,andAndrewWhitten.2021. “IPOsandCorporateTaxes.” WorkingPaper . Dowd, Tim, Christopher Giosa, and Thomas Willingham. 2020. “Corporate behavioral responses totheTCJAfortaxyears2017–2018.” NationalTaxJournal73(4):1109–1134. Dyreng, Scott, Fabio B. Gaertner, Jeffrey L. Hoopes, and Mary Vernon. Forthcoming. “The effect of US tax reform on the tax burdens of US domestic and multinational corporations.” ContemporaryAccountingResearch. Dyreng,ScottD.,MichelleHanlon,EdwardL.Maydew,andJacobR.Thornock.2017. “Changes in corporate effective tax rates over the past 25 years.” Journal of Financial Economics 124(3):441–463. 19

Fox, Edward and Benjamin Pyle. 2022. “Who benefits from corporate tax cuts? Evidence from banksandcreditunionsaroundtheTCJA.” WorkingPaper . Gaertner, Fabio B., Brent Glover, and Oliver Levine. 2021. “A re-examination of firm size and taxes.” AvailableatSSRN3928145. Gaertner, Fabio B., Daniel P. Lynch, and Mary E. Vernon. 2020. “The effects of the Tax Cuts and JobsActof2017ondefinedbenefitpensioncontributions.” ContemporaryAccountingResearch 37(4):1990–2019. Gale,WilliamG.,HilaryGelfond,AaronKrupkin,MarkJ.Mazur,andEricJToder.2019. “Effects oftheTaxCutsandJobsAct: Apreliminaryanalysis.” NationalTaxJournal71(4):589–612. Garcia-Bernardo, Javier, Petr Jansky`, and Gabriel Zucman. 2022. “Did the Tax Cuts and Jobs Act reduce profit shifting by U.S. multinational companies?” Tech. rep., National Bureau of EconomicResearch. Graham, John R., Michelle Hanlon, Terry Shevlin, and Nemit Shroff. 2014. “Incentives for tax planningandavoidance: Evidencefromthefield.” TheAccountingReview89(3):991–1023. Hanlon, Michelle, Jeffrey L. Hoopes, and Joel Slemrod. 2019. “Tax reform made me do it!” Tax PolicyandtheEconomy33(1):33–80. Henry, Erin, George A. Plesko, and Steven Utke. 2018. “Tax policy and organizational form: Assessing the effects of the Tax Cuts and Jobs Act of 2017.” National Tax Journal 71(4):635–660. Henry,ErinandRichardSansing.2018. “Corporatetaxavoidance: Datatruncationandlossfirms.” ReviewofAccountingStudies23(3):1042–1070. Hoopes, Jeffrey L., Patrick Langetieg, Edward L. Maydew, and Michele Mullaney. 2020. “Is tax planningbestdoneinprivate?” AvailableatSSRN3420362. Joint Committee on Taxation. 2017. “Estimated budget effects of the conference agreement for H.R.1,The”TaxCutsandJobsAct”.”. ———.2018. GeneralExplanationofPublicLaw115–97. Kennedy, Patrick J., Christine L. Dobridge, Paul Landefeld, and Jacob Mortenson. 2022. “The efficiency-equity tradeoff of the corporate income tax: Evidence from the Tax Cuts and Jobs Act.” WorkingPaper . Kubick, Thomas R., G. Brandon Lockhart, and John R. Robinson. 2021. “Taxes and earnings management: EvidencefromtheTaxCutsandJobsActof2017.” AvailableatSSRN3891629. 20

Laplante, Stacie Kelley, Christina Lewellen, Dan Lynch, and David M.P. Samuel. 2021. “’Just BEAT It’ Do firms reclassify costs to avoid the base erosion and anti-abuse tax (BEAT) of the TCJA?” WUInternationalTaxationResearchPaperSeries (2021-03). Mills, Lillian F. and Kaye J. Newberry. 2001. “The influence of tax and nontax costs on book-tax reporting differences: Public and private firms.” Journal of the American Taxation Association 23(1):1–19. Ohrn, Eric. 2018. “The effect of corporate taxation on investment and financial policy: Evidence fromtheDPAD.” AmericanEconomicJournal: EconomicPolicy10(2):272–301. Rubin, Richard and Theo Francis. 2018. “Did the U.S. tax overhaul do what it promised?” Wall StreetJournal. ———. 2021. “In House tax bill, companies get return of higher rates but not the breaks.” Wall StreetJournal. Slemrod, Joel. 2004. “Are corporate tax rates, or countries, converging?” Journal of Public Economics88(6):1169–1186. U.S.InternalRevenueService.2013. StatisticsofIncomeCorporationSourceBook. Wagner, Alexander F., Richard J. Zeckhauser, and Alexandre Ziegler. 2020. “The Tax Cuts and JobsAct: Whichfirmswon? Whichlost?” Tech.rep.,NationalBureauofEconomicResearch. Wier, Ludvig S. and Gabriel Zucman. 2022. “Global profit shifting, 1975-2019.” Tech. rep., NationalBureauofEconomicResearch. Zwick, Eric and James Mahon. 2017. “Tax policy and heterogeneous investment behavior.” AmericanEconomicReview107(1):217–248. 21

Online Appendix A: Variable Definitions The table below provides our variable definitions. All data are sourced from the IRS. Forms and linesreferencedarefor2016unlessotherwisenoted. TableA1: Variabledefinitions Variable Definition Taxespaid Form1120: line31 Domestic cash effective When a firm reports attaching a Schedule M-3 to the Form 1120 (Box taxrate(ETR) A4): Taxes paid/[Net income (loss) per income statement of includible corporations (Schedule M-3, Part I, line 11) + U.S. current income tax expense(ScheduleM-3,PartIII,line1)+U.S.deferredincometaxexpense (ScheduleM-3,PartIII,line2)] When a firm does not report attaching a Schedule M-3: Taxes paid/[Net income (loss) per books (Schedule M-1, line 1) + Federal income tax per books(ScheduleM-1,line2)]. Positivetaxespaid Indicatorvariableequalto1ifTaxesPaid > 0 (0/1) Totalassets Form1120: ScheduleL,line15(columnd) Sales Form1120: line1c Depreciableassets Form 4562: sum of lines 19a (column c) to line 19i (column c) + line 20a (column c) + line 20b (column c) + line 20c (column c) + line 14 + line 15 +line16 Netincome Form1120: line28 Taxableincome Form1120: line30 Payroll Sum of total wages reported on employee Form W-2s (Box 5, Medicare Wages) of a given firm, after SOI firm sample is merged to universe of employee W-2s (see Dobridge, Landefeld, and Mortenson (2021) for more detailontheSOImerge) U.S.employment Sum of total firm employees after SOI sample is merged to universe of employee W-2s, as described in Dobridge, Landefeld, and Mortenson (2021) Paidthe Indicator variable equal to 1 if the corporate alternative minimum tax corporateAMT (AMT)(Form1120: ScheduleJ,line3)> 0 HadanNOL Indicator variable equal to 1 if the net operating loss (NOL) stock (Form carryforward 1120: ScheduleK,line12)> 0 22

Variabledefinitions UsedanNOL Indicatorvariableequalto1ifForm1120: line29a> 0 carryforward BonusorHad Indicator variable equal to 1 if a firm reported any capital expenditures on depreciablecapital Form4562. expenditures ClaimedtheDPAD Indicator variable equal to 1 if the Domestic Production Activities Deduction(DPAD)(Form1120: line25)> 0 Hadpositivenet Indicator variable equal to 1 if interest deduction minus interest income interestexpense(0/1) (Form1120: line18minusForm1120: line5)> 0 Aboveinterest Indicator variable equal to 1 if net interest expense (Form 1120: line 18 limitation minus Form 1120: line 5) > 0 30 percent of firm earnings before interest, taxesanddepreciation(EBITDA)(Form1120: line28plusForm1120: line 18+Form1120: line20+Form1120: line25) Aboveinterest Indicator variable equal to 1 if Above interest limitation = 1 and if Taxable limitation,> 0taxable income> 0 income Positivenetinterest Indicatorvariableequalto1ifdepreciationdeduction(Form1120: line18) expense > 0 Unrepatriatedforeign Indicator variable equal to 1 if total foreign earnings and profits (E&P) earnings reportedonForm5471: ScheduleH,line5disnon-zero PositiveforeignE&P Indicator variable equal to 1 if total foreign earnings and profits (E&P) reportedonForm5471: ScheduleH,line5disnon-zero Repatriatedearnings Indicator variable equal to 1 if positive dividends reported on Form 1120: ScheduleC,lines7,8,or13. ForeignE&Pinexcess Indicator variable equal to 1 if foreign E&P (Form 5471: Schedule H, line ofDTIR 5d) is > deemed tangible income return (DTIR). DTIR is 10% of Form 5471ScheduleF:line8a 23

Online Appendix B: Supplementary Figures and Tables FigureB1: Industrydistributionbyfirmtype (a)PublicMultinational (b)PublicDomestic (c)PrivateMultinational (d)PrivateDomestic This figure presents the share of firms in the sample in each SOI industry by firm type. Panels (a), (b), (c), and (d) show the distributions for public multinationals, public domestic firms, private multinationals, and private domestic firms,respectively. AllpanelswerecreatedbytheauthorsusingIRSadministrativetaxdataandvariablesaredefined inAppendixA.AllsamplesareweightedusingSOIweightstoproducefirm-weightedpopulationaverages. Notethat privatedomesticfirmsmakeup98percentofthetotalfirmsinthesampleandaccountforatleast94percentofthe totalfirmsineachindustrycategory. 24

FigureB2: EffectiveTaxRatesforCCorporations: 2010to2020 Overall Public MNE Public Domestic Private MNE Private Domestic Percent 20 18 16 14 12 10 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 (a)ByPublicandMultinationalStatus Overall Bottom 90% 90-95% 95%+ Percent 25 20 15 10 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 (b)BytheDistributionofSales This figure presents trends in average domestic cash effective tax rates from 2010 to 2020 for U.S. C corporations. Panel (a) presents trends for firms overall, for public multinationals (MNEs), public domestic firms, private MNEs andprivatedomesticfirms. Panel(b)showstrendsforfirmsoverall,forfirmsinthe95thpercentileofthefirmsales distribution in 2016, firms in the 90th to 95th percentile of the sales distribution in 2016, and firms below the 90th percentilein2016. AllpanelswerecreatedbytheauthorsusingIRSadministrativetaxdataandvariablesaredefined inAppendixA.Bubblesizesareproportionaltotheaggregateshareof2016federalcorporatetaxrevenuecollected fromeachfirmtype. AllsamplesareweightedusingSOIweightstoproducefirm-weightedpopulationaverages. 25

FigureB3: FractionofU.S.firmspayingU.S.federaltaxes Overall Public MNE Public Domestic Private MNE Private Domestic Share 0.60 0.50 0.40 0.30 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 (a)ByPublicandMultinationalStatus Overall Bottom 90% 90-95% 95%+ Share 0.60 0.50 0.40 0.30 0.20 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 (b)BytheDistributionofSales ThisfigureshowsthefractionofvarioussubsetsofU.S.firmspayingU.S.corporateincometaxfrom2010to2019. Panel (a) shows trends for U.S. firms overall, for public and private multinational firms, and for public and private domesticfirms. Panel(b)showstrendsforfirmsoverall,forfirmsinthe95thpercentileofthefirmsalesdistribution in2016, firmsin the90thto 95thpercentile ofthe salesdistributionin 2016, and firmsbelowthe 90thpercentile in 2016. AllpanelswerecreatedbytheauthorsusingIRSadministrativetaxdataandvariablesaredefinedinAppendix A.Bubblesizesareproportionaltotheaggregateshareof2016federalcorporatetaxrevenuecollectedfromeachfirm type. AllsamplesareweightedusingSOIweightstoproducefirm-weightedpopulationaverages. 26

TableB1: TCJAprovisionsandETRchanges (1) (2) ∆ETR ∆ETR <=2ndMTRBracket 3.904*** 4.083*** (1.037) (1.063) >=4thMTRBracket -9.269*** -9.185*** (1.003) (1.038) AMT -1.625 -2.162** (1.077) (1.092) NOL 4.543*** 4.831*** (0.601) (0.631) Bonus(Inv>0) 0.203 0.130 (0.556) (0.564) DPAD 0.903 0.916 (0.728) (0.861) InterestLimitation -2.074 -2.170 (1.377) (1.351) Multinational -2.060*** -1.659** (0.693) (0.796) Constant -2.425** -2.655** (0.999) (1.042) IndustryFE No Yes R2 0.19 0.21 NFirms 15,889 15,889 This table presents regression results on the association between changes in firms’ ETRs from 2016 to 2019 and pre-TCJAexposuretovarioussignificantprovisionsoftheTCJA,estimatedaccordingtothespecificationinEquation 1. Column (1) presents results without fixed effects (as shown in Figure 3) and column (2) includes two-digit SOI industryfixedeffects(equivalenttotwo-digitNAICSindustries). TheoutcomevariableΔ𝐸𝑇𝑅isthe2016to2019 changeinthecashETR,measuredatthefirmlevel,andalloftheregressorsaredefinedusingfirmcharacteristicsin 2016. ≤1𝑠𝑡𝑀𝑇𝑅𝐵𝑟𝑎𝑐𝑘𝑒𝑡isanindicatorvariableequalto1ifafirmwasinthefirst(15%)taxbracketorpaidzero tax as defined in Table 2; ≥ 3𝑟𝑑𝑀𝑇𝑅𝐵𝑟𝑎𝑐𝑘𝑒𝑡 is an indicator variable equal to 1 if a firm was in the third tax tax bracket or higher as defined in Table 2; 𝐴𝑀𝑇 is an indicator variable equal to 1 if a firm paid the corporate AMT, 𝑁𝑂𝐿isanindicatorvariableequalto1ifafirmusedanNOLcarryforward;𝐵𝑜𝑛𝑢𝑠isanindicatorvariableequalto1 ifafirmreportedanycapitalexpenditureonForm4562;𝐷𝑃𝐴𝐷isanindicatorequalto1ifafirmclaimedtheDPAD; 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝐿𝑖𝑚𝑖𝑡𝑎𝑡𝑖𝑜𝑛isadummyvariableequalto1ifafirmisabovetheinterestlimitationandhadpositivetaxable income;and𝑀𝑁𝐶 isadummyvariableequalto1ifafirmhasaCFCduringthesampleperiod. Furtherinformation about variableconstruction isincluded in AppendixA. The regression is weightedusing SOIsampling weights and thedatasourceisIRSadministrativetaxdata. Standarderrorsareclusteredbyfirmandreportedinparentheses. ***, **and*indicatelevelsof1percent,5percent,and10percentsignificance,respectively. 27