effects of ifrs adoption on tax-induced incentives for financial earnings management: evidence from...

Available online at www.sciencedirect.com

The International Journal of Accounting 48 (2013) 218–247

Effects of IFRS Adoption on Tax-induced Incentivesfor Financial Earnings Management:

Evidence from Greece☆

Nikolaos I. Karampinis, Dimosthenis L. Hevas ⁎

Department of Accounting and Finance, Athens University of Economics and Business, Greece

Received 24 May 2011

Abstract

We investigate whether the adoption of International Financial Reporting Standards (IFRS) inGreece affected tax-induced incentives for financial earnings management. Prior to the imple-mentation of IFRS, there were powerful incentives for firms facing higher tax pressure to restrict(exacerbate) upward (downward) financial earnings management due to direct tax implications. IFRSadoption reduced book–tax conformity, thereby releasing financial income from tax implications. Asexpected, we find that tax pressure is a significant negative determinant of discretionary accruals inthe pre-IFRS period. However, this effect dissipates under the new IFRS regime.© 2013 University of Illinois. All rights reserved.

JEL classification: M41; H26Keywords: Book–tax conformity; IFRS adoption; Earnings management; Greece

1. Introduction

Using a sample of Greek firms, we study the potential implications of the adoption ofInternational Financial Reporting Standards (IFRS henceforth) for book–tax conformityand its concomitant effects on tax-induced managerial opportunism. Greece provides aninteresting research setting because high book–tax conformity and close links between

☆ We appreciate the helpful comments from Thomas Omer (the editor), two anonymous referees, Dimitrios Ghicas,Apostolos Ballas, and participants at the 2011 European Accounting Association Conference.⁎ Corresponding author.

E-mail address: [email protected] (D.L. Hevas).

0020-7063/$ - see front matter © 2013 University of Illinois. All rights reserved.http://dx.doi.org/10.1016/j.intacc.2013.04.003

mailto:[email protected]

http://dx.doi.org/10.1016/j.intacc.2013.04.003




219N.I. Karampinis, D.L. Hevas / The International Journal of Accounting 48 (2013) 218–247

financial income and taxable income prevailed prior to IFRS enactment. The stronginterconnectedness between financial and tax reporting amplifies tax-induced incentives torestrict (exacerbate) upward (downward) financial earnings management for tax purposes.Tax-induced incentives are further encouraged by the paucity of analyst coverage (Chang,Khanna, & Palepu, 2000) and the underdevelopment of the Greek capital market relative toother European jurisdictions (La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 1997). Thelack of analysts' benchmarks to meet or beat, the concentrated family-ownership of Greekfirms (La Porta et al., 1997; Papas, 1992), and the low reputation costs for managers promotetax goals relative to other managerial targets.1 In addition, corporate costs of financial incomemanipulation are not severe due to poor monitoring mechanisms at the institutional level.Weak legal enforcement, low regulatory quality and inadequate shareholder protection areprevalent attributes of the Greek setting (Karampinis & Hevas, 2011).

High book–tax conformity renders upward financial earnings management a particularlycostly activity; even artificially increased financial income entails tax implications. In asimilar vein, high book–tax conformity renders downward financial earnings management apotentially attractive vehicle to reduce taxes. However, these incentives are not expected to beequal across firms. On the contrary, we predict that their intensity varies depending on afirm's tax pressure. We use the term “tax pressure” to refer to a firm's ability to retain lowtaxes relative to its operating performance. Firms facing higher tax pressure exploit lessefficiently the provisions stipulated in Greek tax law (e.g., investments in profitableassociates, tax loss carryforwards, tax credits and untaxed reserves, domiciling in areas withlower tax rates) and incur increased tax outlays. Therefore, these firms may have greaterincentives to restrict (exacerbate) upward (downward) financial earnings management. Inthese respects, the first prediction of this paper is that tax pressure constitutes a significantnegative economic determinant of financial earnings management under a high book–taxconformity regime such as the pre-IFRS period in Greece.

The adoption of IFRS caused an inevitable change in the interconnectedness betweenfinancial accounting and tax accounting in Greece. Because IFRS is independent of taxconsiderations (Hung & Subramanyam, 2007), the tax implications of financial incomeweakened considerably in the post-IFRS period. We argue that such relaxation in book–taxconformity renders upward financial earnings management less costly and downwardearnings management less attractive for tax purposes. Therefore, the second prediction ofthis paper is that the aforementioned effect of tax pressure declines in the post-IFRS period.

To assess these predictions, we gauge financial earnings management via abnormaldiscretionary accruals estimated using a Modified Jones model (Dechow, Sloan, &Sweeney, 1995) that includes return on assets (ROA) to control for operating performance(Kothari, Leone, & Wasley, 2005). In addition, we consider positive and negativediscretionary accruals separately. Following prior research (e.g., Gupta & Newberry, 1997;Othman & Zeghal, 2006; Stickney & McGee, 1982; Zimmerman, 1983), we use eachfirm's average effective tax rate (ETR) to measure tax pressure. In particular, we employthe ratio of the current tax expense to operating cash flows (ETRCFO) as an indicator of tax

1 Interestingly, Wang (2006), using a U.S. sample, finds that family ownership is negatively related to earningsmanagement activities. However, the opposite appears to hold in jurisdictions characterized by weak investorprotection, low-quality GAAPs and private channels of information (Fan & Wong, 2002).

220 N.I. Karampinis, D.L. Hevas / The International Journal of Accounting 48 (2013) 218–247

pressure; higher (lower) values of ETRCFO indicate firms with relatively higher (lower) taxpressure. According to our predictions, in the pre-IFRS period, higher tax pressureintensifies tax-induced incentives either to restrict aggressive financial reporting or tomanipulate financial income downward due to high book–tax conformity.2 Furthermore,relaxations in book–tax conformity as imposed by IFRS implementation release financialaccounting income from direct tax implications and attenuate the effect of tax pressure.

The empirical results confirm that tax pressure exhibits a significant negative relationshipwith earnings management in the pre-IFRS period, regardless of the sign of discretionaryaccruals. This finding indicates that higher tax pressure restricts (exacerbates) upward(downward) financial earnings management. In addition, we find that IFRS adoption had anincremental positive impact on tax pressure, thereby attenuating its negative relationship withdiscretionary accruals. At the aggregate level, we report preliminary evidence that financialincome is recognized in a more aggressivemanner relative to taxable income in the post-IFRSperiod and thus that higher book–tax differences emerge. These results remain robust acrosssensitivity tests concerning the assumed discretionary accruals model and across alternativeexplanations. Overall, these findings suggest that high book–tax conformity and lowcorporate costs induce firms with higher tax pressure to manipulate their discretionaryaccruals accordingly. In the post-IFRS period, the reduction in book–tax conformity rendersfinancial income independent from tax implications, leading this effect to dissipate.

This paper contributes to the literature in two ways. First, we provide evidence on theconsequences of changes in book–tax conformity in a financial reporting system withinadequate monitoring mechanisms. Notice that the existing empirical evidence on theeffects of different levels in book–tax conformity (e.g., Ali & Hwang, 2000; Atwood,Drake, & Myers, 2010) is still limited. In addition, Guenther, Maydew, and Nutter (1997)and Hanlon, Maydew, and Shevlin (2008) are the only studies investigating changes inbook–tax conformity in the U.S. Examining the effects of changes in book–tax conformityin a country-specific sample, rather than stable relationships in multinational samples,ensures that the rest of the institutional characteristics remain constant and provides a morepowerful research setting. Furthermore, empirical research in economic contexts other thanthe U.S. provides new insights in the book–tax conformity literature. These inferences addto the current debate concerning pros and cons of increased book–tax conformity (seeAtwood et al., 2010; Desai, 2005; Hanlon & Shevlin, 2005).

Second, we contribute to a rapidly growing literature concerning the impacts ofIFRS adoption on several aspects of the financial reporting system, such as earningsinformativeness (Landsman, Maydew, & Thornock, 2011), earnings persistence (Atwood,Drake, Myers, & Myers, 2011), cost of capital (Daske, Hail, Leuz, & Verdi, 2008; Li,2010), and analysts' forecast accuracy (Byard, Li, & Yu, 2011). To the best of ourknowledge, this is the first study in the IFRS literature that explores IFRS effects on book–tax conformity and tax-induced incentives for financial earnings management. Ourfindings suggest that a greater separation of taxable income and financial accountingincome, as imposed by IFRS adoption, amplifies financial reporting aggressiveness injurisdictions with high pre-existing book–tax conformity and low-quality monitoring

2 Aggressive financial reporting refers to practices that boost financial income or incorporate positive economicincome to full extent.


mechanisms. These inferences may be extrapolated to other jurisdictions with similarcharacteristics that also adopt IFRS and are concerned about earnings management. Inthese respects, our study extends the evidence reported in Jeanjean and Stolowy (2008) thatearnings management increased in France following IFRS adoption.

The rest of the study proceeds as follows. Section 2 reviews past literature relevant to thisstudy. Section 3 describes book–tax conformity in the Greek setting and the changes thatoccurred with IFRS adoption. Section 4 delineates our research design and presents the testedhypotheses. Empirical results are reported and discussed in Sections 5 and 6. Section 7addresses some additional concerns. Finally, Section 8 summarizes and concludes.

2. Literature review

We perceive two interrelated strands of past research as relevant to our study. The firststrand provides compelling empirical evidence that tax implications trigger financialearnings management. The second strand explores the effects of book–tax conformity onfinancial reporting. We discuss each strand in turn.

Several studies indicate that managers opportunistically manipulated financial incomearound the period of the TaxReformAct in 1986 (TRA86, henceforth) in theU.S. In particular,Scholes, Wilson, and Wolfson (1992), Guenther (1994), Maydew (1997), and Lopez, Regier,and Lee (1998) find evidence consistent with financial income shifting in response toreductions in statutory tax rates. Furthermore, Boynton, Dobbins, and Plesko (1992), Dhaliwaland Wang (1992), and Wang (1994) report that firms affected by the Alternative MinimumTax rule, as imposed by TRA86, shifted income across periods to reduce the tax impact. In astudy more closely related to ours, Guenther et al. (1997) examine a sample of firms that wereforced to switch from the cash method to the accrual method for tax purposes. They find thatfirms tend to defer their financial income after the switch, i.e., increased book–tax conformitycaused financial income manipulation. In a similar vein, Roubi and Richardson (1998) findevidence of earnings management in response to reductions in corporate tax rates in Canadaand Singapore but weaker results for Malaysia. Finally, Lin (2006) detects earningsmanagement in response to schedules of tax rate increases in China. In summary, managersappear to engage in financial income management when the tax implications are material.

Turning to book–tax conformity, the relevant studies attempt to detect potential effectson financial income. Using the same sample in Guenther et al. (1997), Hanlon et al. (2008)find that the increased book–tax conformity caused a decrease in earnings informativeness.However, because the U.S. financial reporting regime is not closely linked to taxation(Hanlon et al., 2008), most studies directly or indirectly explore book–tax conformityeffects using international samples.3 For instance, Joos and Lang (1994), Ali and Hwang(2000), and Hung (2000) examine book–tax conformity effects on the value relevance ofearnings. Guenther and Young (2000) explore the potential effects on the relationship of

3 A limited number of international studies provide direct evidence on book–tax conformity effects usingrelevant indexes (e.g., Hung, 2001; Atwood et al., 2010), whereas the majority of the literature perceives book–taxconformity as part of the overall status of the country under examination (e.g., Joos & Lang, 1994; Lara et al.,2005). Watts (2003), in his thorough discussion on accounting conservatism, pinpoints the lack of studies directlyinvestigating the effects of taxation. This lack is notable because taxation has always been considered a criticalfactor of accounting differences worldwide (Eberhartinger, 1999; Nobes & Parker, 2008).


earnings with economic activity. Lara, Osma, and Mora (2005) and Bushman and Piotroski(2006) are concerned with potential impacts on conditional conservatism, whereas Lara etal. (2005), Coppens and Peek (2005), and Othman and Zeghal (2006) (2006) seekimplications on earnings management. Finally, Atwood et al. (2010) explore book–taxconformity consequences regarding earnings persistence and predictability. In general,empirical results of the above international studies are often mixed and difficult tointerpret, which is likely due to different samples or time periods, different book–taxconformity measures, and other correlated (omitted or not) institutional variables.

However, additional empirical evidence on book–tax conformity effects is needed. Thepotential benefits and costs of high book–tax conformity have been debated in the U.S. overfive years (Atwood et al., 2010). In general, proponents assert that increased book–taxconformity restricts aggressive financial reporting and improves earnings quality (Desai,2005). Opponents argue that the financial reporting system and the tax reporting systempursue different economic objectives. In particular, financial reporting is intended to provideoutside stakeholders (e.g., investors, creditors, regulators) with information that is useful inmaking financial decisions. On the other hand, tax reporting intends to ensure governmentrevenues, redistribute wealth, achieve equity and social objectives, encourage certaineconomic activities and achieve macroeconomic objectives (Hanlon, Laplante, & Shevlin,2005). Therefore, increased book–tax conformity would cause an efficiency loss in one orboth of the two systems (Hanlon & Shevlin, 2005; Hanlon et al., 2005; Hanlon, Maydew, &Shevlin, 2006; Hanlon et al., 2008).

We attempt to detect the effects of book–tax conformity on earnings management in apowerful research setting. In particular, we examine book–tax conformity changes inGreece, which is typically classified as a continental code-law country with high book–taxconformity (Karampinis & Hevas, 2011; La Porta et al., 1997). We chose Greece for severalreasons. First, unlike common-law countries (e.g., the United Kingdom, the U.S., andAustralia), where there is independence between taxation and financial reporting (Porcano& Tran, 1998), in continental Europe, the two systems have occasionally coincided.Unsurprisingly, tax objectives traditionally dominated in the state-supervised standardsetting process.4 Therefore, domestic accounting standards have been endogenouslydeveloped to meet tax objectives, and the tax implications of financial income are direct.

Furthermore, tax motivations are likely to be more attractive in markets where corporatecosts are of second-order importance. Firms' reliance on bank institutions for financialneeds, the underdevelopment of capital markets (La Porta et al., 1997), the paucity ofanalysts' benchmarks (Chang et al., 2000), and firms' concentrated ownership in code-lawjurisdictions (Rajan and Zingales, 2003) prioritize tax goals in managerial activities. Priorevidence implies that low corporate costs and high book–tax conformity in continentalcountries induce managers to restrict aggressive financial reporting and manipulate incomedownward. Furthermore, minimum earnings management thresholds are determined bydebt covenants, compensation plans and stable dividend streams, rather than meeting or

4 The level of book–tax conformity is clearly related to the degree of a state's intervention in the formation ofaccounting standards (see, for instance, Table 2 in Ali and Hwang (2000)). Because the establishment ofaccounting standards in most continental code-law countries was assigned to governmental standard-settingbodies, these jurisdictions are usually characterized by high book–tax conformity.


beating earnings forecasts (Kasanen, Kinnunen, & Niskanen, 1996; Lara et al., 2005).Therefore, the likelihood of detecting tax-induced incentives in such research contexts ishigher than in common-law settings.

Finally, a release of financial reporting from taxation in a country-specific sample mayaffect these tax-induced opportunistic propensities while the rest of the institutionalvariables remain constant. Therefore, the potential effects are directly attributed toreductions in book–tax conformity rather than to other institutional changes. In theserespects, we argue that the Greek financial reporting system provides an interestingresearch setting to examine how changes in book–tax conformity affects managerialopportunism. In particular, in the pre-IFRS period, Greek Accounting Standards required ahigh degree of alignment between financial and tax concepts, whereas corporate costs werearguably low. In the post-IFRS period, corporate costs remain at similar levels, but the linkbetween tax and financial accounting has materially weakened. The next section discussesthe influence of taxation on the development of the Greek financial reporting system andthe impact of IFRS adoption in greater depth.

3. Book–tax conformity in Greece

Since 1949, taxation has played a major role in the development of the Greek financialreporting rules. In 1949, Greece was at the tail end of a series of wars, including wars against theOttoman Empire, BalkanWars, World War I and II, and the Greek Civil War. The aftermath ofthese combats was the absolute devastation of the country's infrastructures, political upheavaland turmoil, and economic stagnation. Accounting regulation, which was first introduced by theCorporate Law 2190/1920, had remained underdeveloped, while there were few ramificationsfor financial noncompliance. In an ostensible effort to collect necessary financial resources forthe State's restructuring, the Greek government, under the instigation of the American Missionto Aid Greece (AMAG), promulgated the first Tax Code in 1947 (Ballas & Hevas, 2000).5 TheTax Code prescribed certain rules of income determination and heavy penalties in the cases oftax evasion, which culminated in its massive adoption by Greek entities. Due to the absence ofofficial Greek Accounting Standards, Greek entities adopted the Tax Code to serve as the basisfor documenting their economic transactions. Therefore, prior to the establishment of the GreekGeneral Accounting Plan (GGAP), financial reporting rules followed the Greek tax legislation,which was accomplished through various legal forms, such as laws, legislative decrees,compulsory laws, royal and presidential decrees, and subsequent Tax Codes. However, thesemeasures were intended to determine the calculation process of taxable income and ensure thestate's proceeds from taxes instead of measuring firms' real economic performance.

The GGAP was developed in 1980 and became compulsory in 1991 in an attempt toharmonize the valuation methods and bookkeeping of Greek entities. Remarkably, thedevelopers of the GGAP strove to incorporate the guidelines required by the 4th and 7thEuropean Directives.6 However, because the Greek state retained an active role in the

5 The actual motivation was apparently the gradual admission of the Western-supported Greek Governmentagainst guerrilla forces during the Greek Civil War.6 For the first (and the last) time in Greek accounting history, the committee responsible of the preparation of

GGAP was an assemblage of academics, government officials, accounting practitioners and representatives of theBody of Greek Certified Accountants and the Federation of Greek Industries.


standard setting process, the GGAP was to a large extent codified to improve the taxcollection procedure and facilitate tax inspections. In addition, several valuation methodsimposed by the Tax Code were also adopted by the GGAP, whereas taxable and financialincome remained strongly interrelated (Eberhartinger, 1999; European Commission, 1998;Hevas, 1995). Not surprisingly, in accounting areas where the Greek tax legislation andGGAP diverged, Greek accountants followed the tax rules. The adherence to tax rules foraccounting valuation and measurement methods was further justified by the severepenalties for deviation from the orderly bookkeeping, which underscored the tax role offinancial statements.

Panel A of Table 1 presents major accounting areas where the Greek tax legislationoverrides GGAP. This table also reports the respective accounting rules imposed byIFRS. As stated, tax legislation overrides GGAP in critical accounting areas, such asdepreciation, allowance for doubtful accounts, and post-retirement benefits. Notsurprisingly, accounting rules determined by tax legislation diverge from those imposedby IFRS.

In addition, Panel B of Table 1 presents accounting areas where the Greek taxlegislation and GGAP coincide but diverge to IFRS. The acceptance of alternativeaccounting treatments concerning intangible assets and capitalized costs as well as thepermission of LIFO provides leeway for income smoothing, direct write-off charges andearnings management under GGAP (Karampinis & Hevas, 2011).

In summary, prior to IFRS adoption, book–tax conformity was high in Greece, as theprimary role of financial statements was the calculation of taxable income and itsverification by tax authorities.7 This setting amplifies tax-induced incentives for earningsmanagement, which is rampant in Greece (Caramanis & Lennox, 2008; Leuz, Nanda, &Wysocki, 2003). Moreover, other forces that shape preparers' (managers' and auditors')incentives, such as auditors' reputation and professionalism, analysts' benchmarks,litigation risk, international influences and market pressure, were of secondary importance(Karampinis & Hevas, 2011).

Mandatory IFRS adoption in 2005 caused an important change in the status quo of theGreek reporting system. IFRS are independent of tax considerations (Hung &Subramanyam, 2007) and impose a set of accounting rules that are mainly based on thefair view of accounting transactions unlike the tax orientation in Greek AccountingStandards (GAS).8 The investor orientation in IFRS implies that the pre-existing highbook–tax conformity inevitably weakens (Nobes & Parker, 2008). Consequently, todayGreek listed firms prepare their financial statements under IFRS but also retain a separateset of books and records in accordance with GGAP and tax legislation to accommodate thecalculations of taxable income and tax inspections (Grigorakos, 2006). Presumably, therelease of financial reporting from tax implications attenuates tax-induced incentives forfinancial earnings management.

7 Until recently, the prevailing concept in Greece was that financial statements are primarily prepared to fulfilltax needs. This view has been pinpointed and criticized in several accounting textbooks (e.g., Thanopoulou, 2001,pp. 25, 44, 272–273).8 It is notable that several studies ascertain a wide variance between IFRS and GAS (Ding, Hope, Jeanjean, &

Stolowy, 2007; Karampinis & Hevas, 2011; Nobes, 2001).


4. Main hypotheses and research design

In the pre-IFRS period, high book–tax conformity intensifies tax implications offinancial income. Therefore, financial income constitutes a potential vehicle fortax-induced opportunistic incentives. Tax-induced incentives are expected to be moreintense for firms facing higher tax pressure. We use the term “tax pressure” to refer to theability of a firm to retain low taxes relative to its operating performance. Firms with highertax pressure exploit less efficiently the provisions stipulated in the Greek tax legislation(e.g., investments in profitable associates, tax loss carryforwards, tax credits and untaxedreserves, domiciliation in areas with lower tax rates) and incur increased tax outlays.Whereas the term “tax avoidance” in the sense of Dyreng, Hanlon, and Maydew (2008)may be appropriate for our study, it has also been used to describe fraudulent tax reporting(Hanlon & Heitzman, 2010), which is not necessarily implied here. Furthermore, termssuch as tax aggressiveness usually refer to tax planning strategies that encompass a rangeof activities characterized as tax shelters (e.g., transfer pricing arrangements, the location ofintangible property in tax havens, synthetic lease agreements) and may fall in the gray areaof the tax code or even be illegal (see Chen, Chen, Cheng, & Shevlin, 2010; Frank, Lynch,& Rego, 2009). Whereas Greek firms may exercise many of these practices, we do notintend to capture exclusively the extent to which a firm is associated with these practices.

In these respects, we predict that higher tax pressure enhances tax-induced incentives torestrict aggressive financial reporting or manipulate financial income downward, i.e., anegative relationship between tax pressure and financial earnings management is expected.Thus, our first hypothesis states the following:

H1. In the pre-IFRS period, tax pressure is a significant negative economic determinant offinancial earnings management.

IFRS mandatory adoption reduced book–tax conformity and the tax implications offinancial income. Therefore, we hypothesize that the effect of tax pressure weakens in thepost-IFRS period. Thus, our second hypothesis states the following:

H2. In the post-IFRS period, the effect of tax pressure, as a negative economic determinantof financial earnings management, has declined.

We follow the prior literature and use discretionary accruals to measure financialearnings management. To estimate discretionary accruals, we employ a Modified Jonesmodel (Dechow et al., 1995) including return on assets (ROA) to control for operatingperformance as suggested by Kothari et al. (2005):

ACCit ¼ b0jt þ b1jt 1=ASSETit�1ð Þ þ b2jt ΔREVit−ΔARitð Þ þ b3jtGPPEit þ b4jtROAit þ εit;

ð1Þwhere ACCit is total accruals scaled by lagged total assets for firm i in year t, ASSETit istotal assets for firm i in year t, ΔREVit is change in revenues scaled by lagged total assetsfor firm i in year t, ΔARit is change in accounts receivable scaled by lagged total assets for

Table 1Accounting areas that Greek tax legislation overrides GGAP or diverges to IFRS.Source: Greek Tax Code, Code of Books and Records, Greek General Accounting Plan, International Financial Reporting Standards, Karampinis and Hevas (2011).

Panel A. Major accounting areas that Greek tax legislation overrides GGAP

Accounting area Tax legislation Greek General Accounting Plan IFRS

Depreciation 1. Depreciation rates are predetermined.2. Historically, depreciation charges wereoptional during specific time periods. a

3. Extra depreciation rates are predeterminedby the tax legislation, and charges are allowedduring specific time periods.

1. Depreciation rates are based on theestimated useful life of assets.2. Depreciation charges are compulsoryfor assets in use.

1. Depreciation rates are based on theestimated useful life of assets.2. Depreciation charges are compulsoryfor assets in use.

Allowances for doubtful accounts 1. The upper limit of bad expense in eachperiod is predetermined by specific ratesmultiplied with annual revenues and tradedebtors closing balance.2. Historically, the direct write-off methodwas required during specific time periods. b

1. There is no upper limit on the baddebt expense. Allowances for doubtfulaccounts are based on estimations aboutthe receivables that will be most likelyuncollectible in the subsequent year.

1. Accounts receivable are reported at thelower of the two between historic cost andnet realizable value.

Obligations for post retirementbenefits

They are restricted to compensation ownedto employees that will be retired in thesubsequent year.

Obligations are estimated as a percentagemultiplied by compensation owed wereall employees to retire in the subsequentyear.

The projected benefit valuation methodshall be used.

Investments in stock for tradingpurposes

1. All firms are required to follow the costmethod for subsequent valuation except forlisted firms, which are allowed to use marketprices at the closing date.2. Loss occurred from trading or subsequentmeasurement is recognized directly in equityaccounts.3. Profits from trading listed shares aretax-exempt provided that they are recognizeddirectly in equity accounts.

1. All firms are required to follow thecost method for subsequent valuation.2. Loss occurred from trading orsubsequent valuation is recognized in theincome statement.3. Profits from trading shall be recognizedin the income statement.

1. The fair value method shall be usedfor subsequent valuation.2. Profits or losses from trading orsubsequent valuation shall be recognizedin the income statement.

226N.I.

Karam

pinis,D.L.H

evas/The

InternationalJournal

ofAccounting

48(2013)

218–247

Tangible assets 1. Assets acquired at a cost lower than specificamounts predetermined by tax legislation canbe expensed in the acquisition year.2. Impairment is not recognized for taxpurposes.3. Revaluation rates are predetermined by taxlegislation and revaluation is compulsory inspecific years.

1. Assets shall be depreciated over theiruseful life.2. If relevant indications exist, impairmentshall be recognized.3. No specific rules are prescribed for assetrevaluation.

1. Assets shall be depreciated over theiruseful life.2. Impairment shall be recognized if therecoverable value is lower than thecarrying value.3. Revaluation is permitted as long as anentity follows the revaluation model orreverses an impairment loss recognizedin prior periods.

Lease contract Capitalization of financial leases is notpermitted.

No specific rules are prescribed. Capitalization of financial leases isrequired when specific criteria are met.

Panel B. Major accounting areas that Greek tax legislation and GGAP coincide but diverge to IFRS

Greek tax legislation and GGAP IFRS

Intangible assets Start-up costs, R&D costs, interest on loans during construction periods, andgoodwill are allowed to be expensed in the year they occur or capitalizedand amortized in a maximum period of five years; for R&D costs inparticular, the maximum amortization period is three years.

1. Start-up and research costs are expensed.2. Development costs are capitalized provided that specific criteria are met.3. Borrowing costs directly attributed to a qualifying asset are included inthe cost of that asset.

Inventory 1. LIFO is permitted.2. Subsequent measurement at the lower of the two between the historiccost and the current replacement cost. The use of the net realizable valueis permitted only if it is lower than the current replacement cost; the lattershould be lower than the historic cost.

1. LIFO is not permitted.2. Subsequent measurement is made at the lower of the two betweenhistoric cost and net realizable value.

Investment inbonds

Bond investments are reported using the straight line method. Bond investments are reported using the effective interest rate method.

Income taxes Only current taxes are reported Deferred tax assets and liabilities shall be recognized for temporarybook-tax differences.

a Depreciation charges were optional during the period 1992 to 1997.b The direct write-off method was required during the period 1955 to 1992.

227N.I.K

arampinis,D

.L.Hevas

/TheInternationalJournalofA

ccounting48

(2013)218–247


firm i in year t, GPPEit is gross property plant and equipment scaled by lagged total assetsfor firm i in year t, and ROAit is pre-tax income scaled by lagged total assets for firm i inyear t. Eq. (1) is estimated in each industry–year jt that has at least six observations(DeFond & Jiambalvo, 1994; Subramanyam, 1996). Discretionary accruals (DACC) arethe residuals of Eq. (1). Because tax pressure may affect upward and downward earningsmanagement differently, we distinguish between positive and negative discretionaryaccruals (DACC+ and DACC−, respectively).

Following prior studies (Gupta & Newberry, 1997; Othman & Zeghal, 2006;Zimmerman, 1983), we measure the tax pressure at the firm level as each firm's effectivetax rate (ETR).9 Therefore, we employ the ratio proposed by Zimmerman (1983):

ETRCFO ¼ Current TaxOperating Cash Flows

:

We rely on ETRCFO rather than other ETRs that use total taxes as the numerator orscale taxes with pre-tax book income for several reasons. First, we use current taxes insteadof total taxes because GAS does not permit deferred taxation. Furthermore, there iscompelling empirical evidence suggesting that deferred taxes are used for or reflectearnings management activities (Cook, Huston, & Omer, 2008; Hanlon, 2005; Phillips,Pincus, & Rego, 2003; Phillips, Pincus, Rego, & Wan, 2004). Hanlon (2003) notes thatcurrent taxes reported by U.S. firms may be over- or understated due to the inclusion of taxdeductions related to exercised employees' stock options and potential contingencyreserves.10 However, none of these items affects current taxes under GAS or IFRS duringour sample period. Second, we compare different accounting regimes that determinefinancial reporting outcomes. Therefore, ETRs that use pre-tax book income as thedenominator are likely to be inappropriate in this case. However, neither current taxes noroperating cash flows are influenced by alternative accounting standards, and ETRCFOremains consistent across accounting regimes. Third, accounting accruals represent a corecomponent of pre-tax income, thereby increasing the likelihood of mechanical correlations.Finally, effective tax rates that use pre-tax book income are affected by non-conformingearnings management (Hanlon & Heitzman, 2010). Non-conforming positive (negative)accruals increase (decrease) pre-tax income but leave current taxes unaffected. In ourresearch design, this effect would cause a spurious negative relationship, particularly in thepost-IFRS period, when the presence of non-conforming accruals is more pronounced.However, we do not expect such effects on ETRCFO.

9 We use average effective tax rates instead of marginal tax rates for two reasons. First, the empiricalimplementation of marginal tax rates requires simulation procedures involving long time series of firm-specifictaxable income (e.g., Shevlin, 1990; Graham, 1996a, 1996b), which are currently unavailable in our dataset.Second, average effective tax rates are more suitable to our research question; discretionary accruals stem from thecurrent-period income, which is the outcome of a firm's current investments. The current-period income is subjectto contemporaneous tax provisions that are included in the average effective tax rate but may not affect taxes ofmarginal investments (Fullerton, 1984).10 Under the U.S. reporting regime, tax benefits related to exercised employees' stock options are not deductedfrom current taxes; instead, they are credited directly to equity. The contingency reserve is related to potentialcosts that will occur (such as tax penalties and fines) if an aggressive tax position is overturned from tax authoritiesin the future (Hanlon, 2003).


To test our hypotheses, we estimate the following regressions:

DACCþit ¼ c0 þ c1ETRCFOit þ c2SIZEit þ c3LEVit þ c4CURRENTit þ c5CFOit þ c6AUDit þ c7IPOit

þ c8DIFRSit þ c9DIFRSit � ETRCFOit þ c10DIFRSit � SIZEit þ c11DIFRSit � LEVitþ c12DIFRSit � CURRENTit þ c13DIFRSit � CFOit þ c14DIFRSit � AUDit þ c15DIFRSit � IPOitþ Ind:Dummies þ YearDummiesþ uit

ð2Þ

DACC−it ¼ d0 þ d1ETRCFOit þ d2SIZEit þ d3LEVit þ d4CURRENTit þ d5CFOit þ d6AUDit þ d7IPOit

þ d8DIFRSit þ d9DIFRSit � ETRCFOit þ d10DIFRSit � SIZEit þ d11DIFRSit � LEVitþ d12DIFRSit � CURRENTit þ d13DIFRSit � CFOit þ d14DIFRSit � AUDit þ d15DIFRSit � IPOitþ Ind:Dummiesþ YearDummiesþ uit

ð3Þ

where SIZEit is the natural log of total assets for firm i in year t; LEVit is total liabilitiesminus cash scaled by total assets for firm i in year t; CURRENTit is current assets scaled bycurrent liabilities for firm i in year t; CFOit is cash flow from operations scaled by laggedtotal assets for firm i in year t; AUDit is a dummy variable taking the value 1 if firm i isaudited by a Big 4 firm in year t and 0 otherwise; IPOit is a dummy variable taking the value1 if firm i has gone public in year t and 0 otherwise; DIFRS is a dummy variable taking thevalue of 1 in the post-IFRS period and 0 otherwise; Ind:Dummies andYearDummies areindustry and year dummies, respectively, and the remaining variables are defined as above.

SIZE is included to control for the size hypothesis, which states that larger firms are morelikely to depress earnings to avoid political attention (Van Tendeloo & Vanstraelen, 2005;Watts & Zimmerman, 1986). LEV is included to control for the debt-equity hypothesis, whichstates that more leveraged firms are more likely to overstate earnings to avoid debt-covenantviolation (Van Tendeloo & Vanstraelen, 2005; Watts & Zimmerman, 1986). CURRENT isincluded because Butler, Leone, and Willenborg (2004) and Caramanis and Lennox (2008))report a positive relationship between abnormal accruals and liquidity.11 CFO is includedbecause Leuz et al. (2003) report an extreme negative relationship between accruals andoperating cash flows in Greece. A negative relationship may also exist between DACC andoperating cash flows. AUD is included to control for differences in audit effectiveness betweenBig 4 and other international or indigenous auditing firms (Becker, DeFond, Jiambalvo, &Subramanyam, 1998). IPO is included to control for positive discretionary behavior in the yearwhen firms go public (Ghicas, Papadaki, Siougle, & Sougiannis, 2008; Teoh, Wong, & Rao,1998). DIFRS reflects the incremental effect caused by IFRS adoption.

Table 2 summarizes the tax rates during our sample period. Through 2004, the statutorytax rate in Greece was 35%. According to Law 3296/2004, it decreased to 32% in 2005, 29%in 2006, and 25% thereafter. In 2010, it was further reduced to 24%. Because the change instatutory tax rate is constant across firms for a particular year, we include year dummies tomitigate potential estimations effects. However, reduced tax rates may inherently affect taxincentives for earnings management. We address this concern in Section 7.2.

11 Particularly, Butler et al. (2004) report that poor liquidity is positively correlated to going concern auditopinions and negative accruals.

Table 2Tax rates during sample period.Source: Hellenic Ministry of Finance.

2000–2004 2005 2006 2007–2009 2010

35% 32% 29% 25% 24%


We estimate Eqs. (2) and (3) with truncated regressions to address the fact that positive(negative) discretionary accruals are left-truncated (right-truncated) to zero. In addition, weestimate robust standard errors clustered by firm (Petersen, 2009). According to ourhypotheses, we predict that c1, d1 are significantly negative (i.e., tax pressure negativelyaffects discretionary accruals) and c9,d9 are significantly positive (i.e., IFRS attenuated theeffect of tax pressure).

5. Sample and descriptive statistics

Our sample comprises all listed firms in the Athens Stock Exchange from 2000 to 2010,that is, five fiscal years before and six fiscal years after IFRS adoption. We refer to theperiod from 2000 to 2004 as the pre-IFRS period and 2005–2010 as the post-IFRS period.We collect financial data from ICAP Database, the largest financial data provider inGreece, and through a hand-collection procedure.12 For fiscal year 2004, ICAP Databasereports only the IFRS-restated amounts that were announced in 2005. Because ourempirical models require several lagged variables (e.g., lagged total assets, lagged sales),we include the observations in 2005 in our post-IFRS subsample using the restatedamounts in 2004,13 , 14 However, the observations in 2004 are excluded from the pre-IFRSsubsample because their corresponding lagged values (the observations in 2003) arereported under GAS. Firms that were eventually delisted are included to avoid anysurvivorship bias in the results. Firms from the financial and insurance sector are excludedbecause they follow different accounting methods and procedures as well as firms from theshipping industry.15 Firms with negative book value are also excluded. Finally, we restrictour sample to firms that report current taxes and operating cash flows greater than zero.

We collect the necessary financial data from firms' parent statements because Greekfirms are taxed on an unconsolidated basis. Therefore, potential tax motivations forearnings management may be more pronounced in unconsolidated statements rather thantheir consolidated counterparts. Moreover, the Greek Corporate Law 2190/1920 permitsGreek firms to adjust specific tax-related entries in parent statements when they prepareconsolidated statements (e.g., extra depreciation charges, provisions for bad debts). Initialentries usually reflect tax-induced financial earnings management in parent statements.

12 We hand-collect current tax data from the post-IFRS period because ICAP Database provides only the total taxitem, which includes deferred taxes. We also correct for intra-period tax allocation issues (i.e., distribution ofcurrent tax in different parts of the financial statements), such as the categorization of current taxes in discontinuedoperations, provided that this information is available.13 In sensitivity analysis, we exclude observations in 2005, and our results remain unchanged.14 To retain observations in 2000, we also collect the required lagged observations in 1999.15 Firms from the shipping industry are excluded because, under the auspices of Greek tax legislation, theirtaxable income is estimated in accordance with the register tons and the age of each ship owned.


This information is not traceable in consolidated statements. To mitigate potential effectsof outliers on our inferences, all variables are truncated in each period at the top 5% andbottom 5% of their distribution.

Table 3 breaks down our final sample by year and industry. According to table 2, thecomposition of our sample remains relatively stable in time, although there is a decline inthe available observations the last two years. However, there is a clustering of observationswithin specific industries, thereby supporting the inclusion of industry dummies in theempirical equations.

Table 4 summarizes the descriptive statistics of the variables employed in this study.The mean (median) ACC equals 0.014 (0.001) in the pre-IFRS period and increasessubstantially to 0.037 (0.029) after IFRS adoption. This finding is indicative of increasedfinancial reporting aggressiveness, which is likely due to reduced book–tax conformity inthe post-IFRS period. Because DACC are residuals of cross-sectional regressions, themean DACC hovers near zero for both periods (as a percentage of total assets). Consistentwith our expectations, the mean (median) DACCþ equals 0.056 (0.050) in the pre-IFRSperiod and increases to 0.070 (0.062) in the post-IFRS period. The reported t-test(Wilcoxon test) suggests that the difference in means (medians) is statistically significant.This finding indicates that positive discretionary accruals exhibit an increase after IFRSadoption, consistent with reduced effects of tax implications. However, these results do nothold in the case of negative discretionary accruals. Turning to the remaining variables, themean AUD indicates that on average, Big 4 audit firms audit 60% of the sample firms. Themean IPO indicates that fewer firms have gone public in the post-IFRS period.

Table 3Distribution of sample firms by year and industry.

Industry 2000 2001 2002 2003 2005 2006 2007 2008 2009 2010

Food products 23 22 24 23 14 17 13 11 12 13Textiles, clothing and accessories 14 13 11 12 8 7 8 4 5 2Publishing 17 15 10 12 9 9 12 8 9 5Oil and gas 10 11 11 10 8 8 8 8 7 5Building materials 7 7 7 8 7 7 7 6 6 6Metal and industrial goods 21 21 18 17 16 15 16 15 10 9Food retailers 6 6 6 5 5 4 4 3 0 0Apparel retailers 8 7 7 7 5 5 6 3 5 4Durable household products 4 4 2 4 3 3 2 2 2 1Electrical product retailers 13 11 11 9 6 7 8 4 2 3Industrial suppliers 7 8 8 8 5 5 6 5 5 4General retailers 6 6 6 6 4 4 4 3 2 2Utilities 7 7 5 7 5 5 5 5 5 5Technology 14 14 13 12 6 4 7 3 3 3Transportation services 2 3 4 4 3 4 5 2 3 4Heavy construction 13 13 13 13 7 9 9 10 10 11Health care 7 8 8 7 6 6 7 5 6 6Leisure 9 9 7 6 6 6 6 0 5 6Total 188 185 171 170 123 125 133 97 97 89Pre-IFRS period 714Post-IFRS period 664

Notes: Industry classification is based on ICAP industry codes.

Table 4Descriptive statistics.

Pre-IFRS period Post-IFRS period

N Mean Median Std. Dev. Min Max N Mean Median Std. Dev. Min Max

ACC 714 0.014 ⁎⁎⁎ 0.001 ⁎⁎⁎ 0.132 ⁎⁎⁎ −0.254 0.291 664 0.037 0.029 0.559 −0.154 0.441ΔREV 714 0.113 ⁎⁎⁎ 0.064 ⁎⁎⁎ 0.183 ⁎⁎⁎ −0.182 0.607 664 0.061 0.018 0.230 −0.270 0.890ΔREC 714 0.070 ⁎⁎⁎ 0.038 ⁎⁎⁎ 0.138 ⁎⁎⁎ −0.131 0.426 664 −0.296 0.002 1.354 −5.978 0.446GPPE 714 0.439 ⁎⁎⁎ 0.378 0.261 ⁎⁎⁎ 0.115 1.084 664 0.746 0.367 1.366 0.067 6.391ROA 714 0.094 ⁎⁎⁎ 0.074 ⁎⁎⁎ 0.077 ⁎⁎⁎ −0.042 0.281 664 0.074 0.035 0.160 −0.172 0.663DACC 714 0.001 0.000 0.069 ⁎⁎⁎ −0.131 0.130 664 −0.001 −0.001 0.095 −0.192 0.218DACC+ 358 0.056 ⁎⁎⁎ 0.050 ⁎⁎⁎ 0.041 ⁎⁎⁎ 0.000 0.130 312 0.070 0.062 0.070 0.000 0.218DACC− 356 −0.054 ⁎⁎⁎ −0.047 0.041 ⁎⁎⁎ −0.131 0.000 352 −0.068 −0.049 0.060 −0.192 0.000ETRCFO 714 0.364 ⁎⁎⁎ 0.219 ⁎⁎⁎ 0.411 ⁎⁎⁎ 0.018 1.695 664 0.167 0.067 0.242 0.002 0.966ADJETRCFO 714 0.852 0.373 ⁎⁎ 1.181 0.057 4.798 664 0.774 0.315 1.153 0.0138 4.828SIZE 714 17.71 ⁎⁎⁎ 17.587 ⁎⁎⁎ 1.155 ⁎⁎⁎ 15.721 19.968 664 18.251 18.319 1.341 14.832 20.484LEV 714 0.445 ⁎⁎⁎ 0.449 ⁎⁎⁎ 0.262 ⁎⁎⁎ 0.016 0.976 664 1.12 0.510 2.512 0.084 11.439CURRENT 714 1.989 ⁎⁎⁎ 1.648 ⁎⁎ 1.055 0.676 4.794 664 1.697 1.523 1.105 0.623 4.063CFO 714 0.124 ⁎⁎⁎ 0.093 ⁎⁎ 0.096 ⁎⁎⁎ 0.001 0.337 664 0.934 0.120 2.401 0.001 8.799AUD 714 0.647 ⁎⁎⁎ 1.000 0.478 0.000 1.000 664 0.581 1.000 0.494 0.000 1.000IPO 714 0.088 ⁎⁎⁎ 0.000 ⁎⁎⁎ 0.284 ⁎⁎⁎ 0.000 1.000 664 0.008 0.000 0.087 0.000 1.000PTI (in thousands) 714 5703 ⁎⁎⁎ 2521 ⁎⁎⁎ 7928 ⁎⁎⁎ 11 31,263 664 9875 3501 14,239 5 49,625TI (in thousands) 714 5322 ⁎⁎⁎ 2109 ⁎⁎ 9752 ⁎⁎⁎ 94 38,992 664 8316 2842 14,895 5 48,759CT (in thousands) 714 1833 ⁎⁎ 721 ⁎⁎⁎ 3004 ⁎⁎⁎ 35 13,265 664 2227 562 4117 20 17,040TCFO (in thousands) 714 10,907 ⁎⁎⁎ 4036 ⁎⁎⁎ 15,400 ⁎⁎⁎ 5 53,300 664 24,469 10,700 32,252 1 94,300 ⁎

Notes:1. The difference in means is based on t-tests; the difference in medians is based on Wilcoxon tests; the difference in standard deviation is based on F-tests.2. Variable definitions: ACC is total accruals scaled by lagged total assets; ΔREV is the change in revenues scaled by lagged total assets; ΔREC is the change in accounts receivable scaled by lagged total assets; GPPE is grossproperty plant and equipment scaled by lagged total assets; ROA is pre-tax income scaled by total assets; DACC is discretionary accruals; DACC+ is positive discretionary accruals; DACC− is negative discretionary accruals;ETRCFO is current tax scaled by cash flow from operations; ADJETRCFO is current tax divided by the corresponding statutory tax rate of each year and scaled by cash flow from operations; SIZE is the natural log of total assets;LEV is total liabilities minus cash, scaled by total assets; CURRENT is current assets scaled by current liabilities; CFO is cash flow from operations scaled by lagged total assets; AUD is a dummy variable taking the value 1 if afirm is audited by a Big 4 firm and 0 otherwise; IPO is a dummy variable taking the value 1 if a firm goes public in year t and 0 otherwise; PTI is pre-tax income; TI is estimated taxable income; CT is current tax; and TCFO isoperating cash flows.

⁎⁎⁎ Represents significance at the 1% level.⁎⁎ Represents significance at the 5% level.⁎ Represents significance at the 10% level.

232N.I.

Karam

pinis,D.L.H

evas/The

InternationalJournal

ofAccounting

48(2013)

218–247

Table 5Pre-tax income vs. taxable income.

Year Sum_PTI Sum_TI BTD Ratio

2000 1123 956 167 1.172001 1077 944 133 1.142002 1304 1258 46 1.042003 1305 1315 −10 0.992004 1275 1387 −112 0.922005 1462 1118 344 1.312006 1951 1075 876 1.812007 2159 1399 760 1.542008 1690 1392 298 1.212009 1612 1233 379 1.312010 1358 1056 302 1.29

Notes: Sum_PTI is sum of pre-tax income each year in millions of euros; Sum_TI is sum of taxable income eachyear in millions of euros; BTD is book–tax difference in millions of euros; and Ratio is Sum_PTI to Sum_TI.


The variable under investigation, ETRCFO, has a mean value of 0.364 in the pre-IFRSperiod, suggesting that a substantial portion of a firm's operating cash flows is destined fortax outlays. Not surprisingly, the ratio decreases to 0.167 after IFRS adoption, likely due tothe reduction in the statutory tax rate initiated in 2005. To facilitate the comparison of thetwo periods, we gross up current taxes to taxable income; we divide current taxes by thecorresponding statutory tax rate each year. The calculated taxable income is likely a noisyestimate of the real taxable income (Hanlon, 2003) but mitigates the differences caused bythe reduction in statutory tax rates. We then scale the estimated taxable income with

Fig. 1. Pre-tax income vs taxable income.

Fig. 2. Ratio of pre-tax income to taxable income.

image of Fig.�2


operating cash flows. The calculated variable (ADJETRCFO) changes insignificantly inthe post-IFRS period according to the means. In addition, it exhibits almost equal variancebetween the two periods. These findings support the idea that ETRCFO remains consistentacross the two periods and that any observable change is mainly attributed to the reductionin tax rates, which our regression analysis controls for.

Table 5 presents the total pre-tax book income (in millions) and the estimated taxableincome (in millions) by year, the annual differences (in millions) and the ratio of pre-taxincome to taxable income. For exposition purposes, we also report observations in 2004.To illustrate the trend of the book–tax difference during the periods under examination,Fig. 1 plots the aggregate book income and the aggregate estimated taxable income byyear. As depicted, the two measures almost coincide in the pre-IFRS period. However, inthe post-IFRS period, the aggregate book income escalates in contrast to taxable income,consistent with increased financial reporting aggressiveness. Similar inferences are drawnfrom Fig. 2, which plots the ratio of pre-tax income to taxable income.

Table 6 reports pairwise correlations across the variables. Correlations in the pre-IFRSperiod are reported above the diagonal and post-IFRS below. In most cases, the reportedfigures are consistent with theory and prior evidence. In particular, ACC is positively andsignificantly related to ΔREV and ΔREC and negatively to GPPE (Jones, 1991;Subramanyam, 1996). In addition, ACC exhibits a significant negative relationship withCFO (Dechow, 1994) in both periods, although this relationship is more pronounced in thepre-IFRS period. Consistent with prior evidence (Butler et al., 2004; Caramanis & Lennox,2008; Young, 1999), DACC is positively related to LEV.

Moreover, ETRCFO exhibits a negative and significant relationship with DACC in thepre-IFRS period (−0.21). Interestingly, this negative relationship appears to hold regardless ofwhether DACC is positive or negative (the correlation equals −0.22 and −0.18, respectively).This finding indicates that tax pressure restricts upward earnings management and inducesdownward earnings management. Interestingly, the correlations between ETRCFO anddiscretionary accruals weaken after IFRS adoption. This change supports our prediction thatETRCFO induces income manipulation to a lesser extent in the post-IFRS period.

6. Multivariate results

Table 7 presents separately multivariate results concerning positive and negativediscretionary accruals. LEV exhibits a positive relationship with both positive and negativeaccruals, suggesting that more leveraged firms tend to overstate their earnings and restrictdownward earnings management. In addition, DACC+ is positively related to IPO andnegatively related to AUD.However, the effect of AUD is statistically insignificant and declinesin the post-IFRS period.16 Consistent with prior evidence in Greece (Leuz et al., 2003) and inother code-law jurisdictions (Kasanen et al., 1996), both positive and negative discretionaryaccruals are negatively related to CFO, confirming that Greek managers tend to smoothaccounting income. Interestingly, in either case, the effect of CFO is attenuated in the post-IFRSperiod.
16 Caramanis and Lennox (2008) find a similarly insignificant relationship between positive discretionaryaccruals and a Big 4 auditor. This finding is consistent with Francis and Wang (2008), who report that the presenceof a Big 4 auditor restricts discretionary accruals only in jurisdictions characterized by strong investor protection.

Table 6Correlation matrix.

ACC ΔREV ΔREC GPPE ROA DACC DACC+ DACC− ETRECFO SIZE LEV CURRENT CFO AUD IPO

ACC 0.29*** 0.70*** -0.15*** 0.16*** 0.61*** 0.46*** 0.39*** -0.25*** -0.16*** 0.27*** 0.09** -0.91*** -0.02 0.16***0.44*** 0.55*** 0.13*** 0.46*** 0.25*** 0.17*** 0.04 -0.06 -0.13*** 0.37*** -0.22*** -0.11*** -0.07* 0.09**

ΔREC 0.91*** -0.47*** 0.13*** 0.40*** 0.37*** 0.25*** 0.15*** -0.17*** -0.18*** 0.38*** -0.03 -0.55*** -0.07* 0.17***GPPE -0.82*** 0.59*** -0.84*** 0.16*** -0.02 -0.06 -0.02 0.04 0.00 0.04 -0.05 0.19*** 0.04 0.14***ROA -0.70*** 0.53*** -0.75*** 0.75*** -0.03 0.01 -0.14** 0.16*** -0.27*** 0.07* 0.01 0.20*** 0.05 0.17***DACC 0.15*** 0.20*** 0.05 -0.02 -0.01 1.00*** 1.00*** -0.21*** 0.02 0.21*** 0.02 -0.60*** 0.01 0.06DACC+ -0.03 0.27*** -0.17*** 0.08 0.18*** 1.00*** n/a -0.22*** -0.03 0.09* 0.05 -0.41*** -0.02 0.08DACC− 0.26*** -0.12** 0.26*** -0.20*** -0.23*** 1.00*** n/a -0.18*** 0.13*** 0.09* -0.05 -0.44*** 0.00 0.00ETRECFO -0.01 -0.04 -0.01 0.00 0.08** -0.11*** -0.12** -0.10** 0.04 -0.17*** 0.08** 0.32*** -0.01 -0.05SIZE -0.03 0.02 0.00 0.04 0.07* -0.05 -0.18*** 0.14** 0.08* 0.02 -0.13*** 0.05 -0.01 -0.11***LEV -0.89*** 0.57*** -0.92*** 0.89*** 0.75*** 0.41*** 0.14** -0.21*** -0.03 0.04 -0.64*** -0.25*** -0.11*** -0.03CURRENT -0.02 -0.02 -0.03 -0.01 0.08** 0.02 -0.02 0.02 0.07* 0.16* -0.05 -0.08** 0.09** 0.14***CFO -0.78*** 0.45*** -0.94*** 0.81*** 0.76*** -0.13*** 0.07 -0.26*** 0.01 0.03 0.89* 0.04 0.05 -0.12***AUD 0.09** -0.03 0.10 -0.09** -0.05 -0.07* -0.10* 0.05 0.06 -0.04 -0.08** 0.07* -0.09** 0.01IPO 0.03 0.04 0.03 -0.03 0.04 -0.01 -0.07 0.01 -0.11** 0.04 -0.02 0.01 -0.02 -0.03

ΔREV

Notes:1. Pairwise correlations in the pre-IFRS (post-IFRS) period are reported above (below) the diagonal in the gray (white) area.2. Variable definitions: ACC is total accruals scaled by lagged total assets; ΔREV is the change in revenues scaled by lagged total assets; ΔREC is the change in accountsreceivable scaled by lagged total assets; is gross property plant and equipment scaled by lagged total assets; ROA is pre-tax income scaled by total assets; DACC isdiscretionary accruals; DACC + is positive discretionary accruals; DACC− is negative discretionary accruals; ETRCFO is current tax scaled by cash flow from operations;SIZE is the natural log of total assets; LEV is total liabilities minus cash, scaled by total assets; CURRENT is current assets scaled by current liabilities; CFO is cash flowfrom operations scaled by lagged total assets; AUD is a dummy variable taking the value 1 if a firm is audited by a Big 4 firm and 0 otherwise; and IPO is a dummy variabletaking the value 1 if a firm goes public in year t and 0 otherwise.⁎⁎⁎Represents significance at the 1% level.⁎⁎Represents significance at the 5% level.⁎Represents significance at the 10% level.

235N.I.K

arampinis,D

.L.Hevas

/TheInternationalJournalofA

ccounting48

(2013)218–247

Unlabelled image


Turning to the variable under examination, Table 7 reveals that in the pre-IFRS period,ETRCFO is a statistically significant negative economic determinant of positive andnegative discretionary accruals; the coefficient on ETRCFO,c1 (d1) takes a value of −0.088(−0.071) and is statistically significant at the 1% level.17 This finding indicates that highertax pressure constrains aggressive financial reporting and exacerbates downward earningsmanagement, consistent with H1. This evidence is attributed to high book–tax conformity,which renders upward earnings management a costly activity and downward earningsmanagement attractive with regard to the tax implications.

Unlike in the pre-IFRS period, the coefficient on ETRCFO decreases (in absolute terms)after IFRS adoption. More specifically, the incremental coefficient, c9 (d9), takes a value of0.044 (0.036) and is statistically significant at the 1% (5%) level. These results indicate thatthe effect of tax pressure is attenuated under the IFRS regime, consistent with H2. Weattribute this finding to relaxations in book–tax conformity after the adoption of IFRS.

To summarize, our empirical findings suggest that tax pressure restricts (exacerbates)upward (downward) earnings management in the pre-IFRS period due to high book–taxconformity. This effect weakens after IFRS adoption and the concomitant release offinancial income from tax implications.

7. Additional analysis and robustness tests

7.1. Alternative specifications

We employ alternative specifications of Eq. (1) because several studies draw attentionto estimation problems in earnings management models (inter alia, Dechow et al., 1995;Young, 1999; Bartov, Gul, & Tsui, 2000; McNichols, 2000; Jones, Krishnan, &Melendrez, 2008). In particular, we make the following adjustments: (1) include laggedROA instead of current ROA, a dummy variable indicating the presence of losses, and the12-month cumulative stock return of the prior year as additional control variables (Jones etal., 2008; Kothari et al., 2005; Wang, 2006); (2) estimate abnormal accruals using anunmodified Jones model (Jones, 1991); (3) use working capital accruals instead of totalaccruals (Teoh et al., 1998); and (4) use working capital accruals and a margin model toestimate discretionary accruals (Peasnell, Pope, & Young, 2000).

Our previous inferences remain robust to these adjustments because they cause only trivialchanges to previously reported results. In particular, the inclusion of the three aforementionedcontrol variables does not affect the signs and significance levels of our test variables. Whenwe employ an unmodified Jones model, the magnitude of the coefficient on ETRCFO in thepre-IFRS period is somewhat lower (in absolute terms), but its significance remains at similarlevels. This finding is not surprising, as the Modified Jones model considers change inreceivables to be discretionary. Allowances for bad debt are likely to constitute a convenientvehicle of managing accruals for tax purposes. Our results remain qualitatively similar whenwe use working capital accruals instead of total accruals as well as when we utilize the margin

17 The average total assets is€181 million and the average net income is€8.511 million in the pre-IFRS period.Therefore, when ETRCFO increases by 1%, positive discretionary accruals decrease by an amount equivalent to2% of their net income (−0.088% ∗ 180 / 7.550 = 2%), and negative discretionary accruals decrease by anamount equivalent to 1.7% of their net income (−0.071% ∗ 180 / 7.550 = 1.7%).

Table 7Results relating to discretionary accruals and ETRCFO in the pre- and the post-IFRS period.

Regression model DACCþit ¼ c0 þ c1ETRCFOit þ c2SIZEit þ c3LEVit þ c4CURRENTit þ c5CFOit

þc6AUDit þ c7IPOit þ c8DIFRSit þ c9DIFRSit � ETRCFOit

þc10DIFRSit � SIZEit þ c11DIFRSit � LEVit þ c12DIFRSit � CURRENTitþc13DIFRSit � CFOit þ c14DIFRSit � AUDit þ c15DIFRSit � IPOit

þ Ind:Dummiesþ YearDummiesþ uit

Regression model DACC−it ¼ d0 þ d1ETRCFOit þ d2SIZEit þ d3LEVit þ d4CURRENTit þ d5CFOit

þd6AUDit þ d7IPOit þ d8DIFRSit þ d9DIFRSit � ETRCFOit

þd10DIFRSit � SIZEit þ d11DIFRSit � LEVit þ d12DIFRSit � CURRENTitþd13DIFRSit � CFOit þ d14DIFRSit � AUDit þ d15DIFRSit � IPOit

þInd:Dummiesþ YearDummiesþ uit

Dependent variable DACC+ DACC−

Coef. z-Stat. Coef. z-Stat.

Intercept −0.022 −0.60 −0.047 −1.26ETRCFO −0.088 −3.11 ⁎⁎⁎ −0.071 −2.92 ⁎⁎⁎SIZE 0.002 1.05 0.003 1.65 ⁎

LEV 0.013 4.27 ⁎⁎⁎ 0.017 3.58 ⁎⁎⁎

CURRENT −0.001 −0.43 −0.002 −0.78CFO −0.098 −4.57 ⁎⁎⁎ −0.143 −8.29 ⁎⁎⁎AUD −0.017 −1.59 −0.002 −0.59IPO 0.005 2.79 ⁎⁎⁎ 0.000 0.02DIFRS 0.128 2.12 ⁎⁎ −0.086 −1.69 ⁎DIFRS ∗ ETRCFO 0.044 2.76 ⁎⁎⁎ 0.036 2.38 ⁎⁎

DIFRS ∗ SIZE −0.007 −2.35 ⁎⁎ 0.000 0.10DIFRS ∗ LEV 0.006 0.45 0.020 1.83 ⁎

DIFRS ∗ CURRENT −0.001 −0.31 0.004 1.00DIFRS ∗ CFO 0.087 4.03 ⁎⁎⁎ 0.137 7.94 ⁎⁎⁎

DIFRS ∗ AUD −0.002 −0.31 0.011 1.55DIFRS ∗ IPO −0.031 −1.50 0.016 1.19Ind. Dummies? Yes YesYear Dummies? Yes YesN 670 708Pseudo R2 7.76% 5.73%

Notes:Variable definitions: DACC+ is positive discretionary accruals; DACC− is negative discretionary accruals;ETRCFO is current tax scaled by cash flow from operations; SIZE is the natural log of total assets; LEV is totalliabilities minus cash, scaled by total assets; CURRENT is current assets scaled by current liabilities; CFO is cashflow from operations scaled by lagged total assets; AUD is a dummy variable taking the value 1 if a firm is auditedby a Big 4 firm and 0 otherwise; IPO is a dummy variable taking the value 1 if a firm goes public in year t and 0otherwise; and DIFRS is a dummy variable taking the value 1 in the post-IFRS period and 0 otherwise.⁎⁎⁎ Represents significance at the 1% level.⁎⁎ Represents significance at the 5% level.⁎ Represents significance at the 10% level.


model to estimate discretionary accruals. Regarding the margin model, the statisticalsignificance of the coefficients on ETRCFO is somewhat lower in the pre-IFRS period.However, the estimated coefficients follow a pattern similar to previous analysis. Overall,these adjustments do not appear to affect our earlier inferences.


7.2. Alternative explanations

An alternative explanation for our results may be the reduction in statutory tax ratesinitiated in 2005, the year of IFRS adoption. As previously mentioned, the statutory taxrate in Greece was 35% through 2004. According to Law 3296/2004, it decreased to 32%in 2005, 29% in 2006, and 25% thereafter. The rate was further reduced to 24% in 2010.Whereas the inclusion of year dummies in Eq. (2) and (3) controls for correspondingstatistical effects, our results in the post-IFRS period may be driven by inherently reducedmanagerial incentives to restrict aggressive financial reporting relative to the pre-reductionperiod.18 To address this issue, we create a matched sample of private firms that incurredthe reduction in the statutory tax rate but still report under Greek Accounting Standards. Inthe case that private firms exhibit relatively stable propensities to manipulate accountingincome even after the reduction in statutory tax rates, we enhanced previous inferences.19

To construct our matched sample, we follow a procedure similar to that used by Ball andShivakumar (2005) but we modify it appropriately. In particular, each year we choose asize-matched private firm from the same industry with the listed firm. The matched firm is theclosest in the year-end total assets to the listed firm, provided that they differ by no more than10%. Once a private firm is chosen, it is not considered a potential match for another listedfirm (no replacement). As soon as the construction of the matched sample is accomplished,we conduct an analysis similar to that of listed firms (with IPO excluded from equations).

Table 8 summarizes the estimated results for the matched sample. They reveal thatETRCFO demonstrates a significant negative relationship with upward earningsmanagement; the coefficient on ETRCFO takes a value of −0.091 in the pre-IFRS periodand is statistically significant at the 1% level. In sharp contrast to the listed firms, theincremental effect of ETRCFO in the post-IFRS period is negative, albeit insignificant.These results indicate that despite the simultaneous reduction in statutory tax rates, taxpressure nonetheless restricts upward earnings management. In a similar vein, ETRCFOappears to constitute a negative determinant of downward earnings management; when weuse negative discretionary accruals as the dependent variable, the coefficient on ETRCFOtakes a value of −0.083 and is statistically significant. The incremental effect in thepost-IFRS period is negative but again insignificant. Overall, these findings support ourprior inferences that the diminished effect of tax pressure for the listed firms in thepost-IFRS period emanates mainly from the reduction in book–tax conformity.

18 Prior evidence suggests that changes in statutory tax rates constitute a significant motivation for incomeshifting (e.g., Guenther, 1994; Lin, 2006; Lopez et al., 1998; Maydew, 1997; Roubi & Richardson, 1998; Scholeset al., 1992). Firms' anticipation of reductions in statutory tax rates with the enactment of Law 3296/2004 wouldfurther exacerbate managerial income shifting in the pre-IFRS period. However, the enactment of Law 3296 was apolitical action of the newly elected government in 2004 and was not publicly known in prior years. Becauseobservations in 2004 are not included in our sample, “anticipation effects” do not affect our results in the pre-IFRSperiod.19 We do not compare directly the results for private firms to listed firms because prior literature suggests thatreporting incentives are profoundly influenced by listing status (Ball & Shivakumar, 2005) and that private firmstend to manage income more (Arnedo et al., 2007; Burgstahler, Hail, & Leuz, 2006; Vander Bauwhede et al.,2003). Note, however, that Burgstahler et al. (2006) find that Greece constitutes an exception, with public firmsengaging in earnings management to a greater extent than private firms.

Table 8Results relating to discretionary accruals and ETRCFO in the pre- and the post-IFRS period for the matchedsample of private firms.

Regression model DACCþit ¼ c0 þ c1ETRCFOit þ c2SIZEit þ c3LEVit þ c4CURRENTit þ c5CFOit

þc6AUDit þ c7DIFRSit þ c8DIFRSit � ETRCFOit þ c9DIFRSit � SIZEitþc10DIFRSit � LEVit þ c11DIFRSit � CURRENTit þ c12DIFRSit � CFOit

þc13DIFRSit � AUDit þ Ind:Dummiesþ YearDummiesþ uit

Regression model DACC−it ¼ d0 þ d1ETRCFOit þ d2SIZEit þ d3LEVit þ d4CURRENTit

þd5CFOit þ d6AUDit þ d7DIFRSit þ d8DIFRSit � ETRCFOit

þd9DIFRSit � SIZEit þ d10DIFRSit � LEVit þ d11DIFRSit � CURRENTitþd12DIFRSit � CFOit þ d13DIFRSit � AUDit þ Ind:DummiesþYearDummiesþ uit



Intercept −0.001 −0.02 −0.048 −1.13ETRCFO −0.091 −3.20 ⁎⁎⁎ −0.083 −2.81 ⁎⁎⁎SIZE 0.002 0.59 0.001 0.24LEV 0.019 2.19 ⁎⁎ 0.015 0.02CURRENT 0.007 1.56 −0.001 −0.37CFO −0.147 −6.42 ⁎⁎⁎ −0.123 −6.53 ⁎⁎⁎AUD −0.001 −0.15 0.003 0.61DIFRS 0.031 0.51 0.077 1.33DIFRS ∗ ETRCFO −0.009 −1.06 −0.007 −0.93DIFRS ∗ SIZE −0.001 −0.14 −0.003 −0.90DIFRS ∗ LEV −0.016 −1.50 −0.018 −1.50DIFRS ∗ CURRENT −0.007 −1.34 −0.001 −0.04DIFRS ∗ CFO 0.026 0.89 −0.001 −0.04DIFRS ∗ AUD 0.005 0.64 0.002 0.26Ind. Dummies? Yes YesYear Dummies? Yes YesN 670 708Pseudo R2 8.53% 7.83% ⁎

Notes:Variable definitions: DACC+ is positive discretionary accruals; DACC− is negative discretionary accruals;ETRCFO is current tax scaled by cash flow from operations; SIZE is the natural log of total assets; LEV is totalliabilities minus cash, scaled by total assets; CURRENT is current assets scaled by current liabilities; CFO is cashflow from operations scaled by lagged total assets; AUD is a dummy variable taking the value 1 if a firm is auditedby a Big 4 firm and 0 otherwise; and DIFRS is a dummy variable taking the value 1 in the post-IFRS period and 0otherwise.⁎⁎⁎ Represents significance at the 1% level.⁎⁎ Represents significance at the 5% level.⁎ Represents significance at the 10% level.


7.3. Non-linearity in the relationship between discretionary accruals and tax pressure

Finally, we investigate potential non-linearity in the relation between DACC and taxpressure. Prior evidence suggests that meeting the previous year's earnings and earningssmoothness are crucial targets for firms (Burgstahler & Dichev, 1997), with the latter beingparticularly important to firms domiciled in code-law countries (Arnedo, Lizarraga, &


S nchez, 2007; Kasanen et al., 1996; Vander Bauwhede, Willekens, & Gaeremynck, 2003).We expect similar managerial incentives in the Greek setting because smoothed earnings,inter alia, ensure a stable stream of dividends. Smooth dividend payments are of first-orderimportance to shareholders when the market is not well developed and owners are willing to

Table 9Results relating to discretionary accruals and ETRCFO after controlling for successful firms.

Regression model DACCþit ¼ c0 þ c1DSit þ c2ETRCFOit þ c3DSit � ETRCFOit þ c4SIZEit

þc5LEVit þ c6CURRENTit þ c7CFOit þ c8AUDit þ c9IPOit

þc10DIFRSit þ c11DIFRSit � DSit þ c12DIFRSit � ETRCFOit

þc13DIFRSit � DSit � ETRCFOit þ c14DIFRSit � SIZEitþc15DIFRSit � LEVit þ c16DIFRSit � CURRENTit þ c17DIFRSit � CFOi

þc18DIFRSit � AUDit þ c19DIFRSit � IPOit þ Ind:DummiesþYearDummiesþ uit

Regression model DACC−it ¼ c0 þ c1DSit þ c2ETRCFOit þ c3DSit � ETRCFOit þ c4SIZEit þ c5LEVit

þc6CURRENTit þ c7CFOit þ c8AUDit þ c9IPOit þ c10DIFRSitþc11DIFRSit � DSit þ c12DIFRSit � ETRCFOit

þc13DIFRSit � DSit � ETRCFOit þ c14DIFRSit � SIZEitþc15DIFRSit � LEVit þ c16DIFRSit � CURRENTit þ c17DIFRSit � CFOit

þc18DIFRSit � AUDit þ c19DIFRSit � IPOit þ Ind:DummiesþYearDummiesþ uit



Intercept −0.015 −0.40 −0.047 −1.25DS −0.009 −2.87 ⁎⁎⁎ −0.006 −2.54 ⁎⁎ETRCFO −0.021 −2.62 ⁎⁎⁎ −0.018 −1.97 ⁎DS ⁎ETRECFO −0.061 −2.96 ⁎⁎⁎ −0.049 −2.54 ⁎⁎SIZE 0.002 0.81 0.003 1.73 ⁎

LEV 0.020 3.16 ⁎⁎⁎ −0.016 −1.47CURRENT −0.001 −0.27 −0.002 −0.73CFO −0.086 −3.85 ⁎⁎⁎ −0.139 −7.35 ⁎⁎⁎AUD −0.006 −1.53 −0.002 −0.59IPO 0.005 0.81 −0.001 −0.08DIFRS 0.128 2.11 ⁎⁎ −0.079 −1.55DIFRS ∗ DS 0.009 0.97 −0.011 −1.58DIFRS ∗ ETRCFO 0.012 1.66 ⁎ 0.018 0.92DIFRS ∗ DS ∗ ETRCFO 0.038 2.38 ⁎⁎ 0.029 2.05 ⁎⁎

DIFRS ∗ SIZE −0.007 −2.30 ⁎⁎ 0.001 0.07DIFRS ∗ LEV 0.001 0.05 0.019 1.74 ⁎

DIFRS ∗ CURRENT −0.002 −0.39 0.004 1.08DIFRS ∗ CFO 0.075 3.32 ⁎⁎⁎ 0.134 7.01 ⁎⁎⁎

DIFRS ∗ AUD −0.002 −0.22 0.010 1.45DIFRS ∗ IPO −0.033 −1.60 0.009 0.68Ind. Dummies? Yes YesYear Dummies? Yes YesN 670 708Pseudo R2 7.95% 4.92%

t


retain concentrated firm ownership (Kasanen et al., 1996). In our research design, areasonable expectation is that the negative relationship between discretionary accruals and taxpressure is more pronounced for firms with pre-managed earnings equal or higher thanprevious year's earnings; tax pressure is of greater importance to firms that have passed theprior year earnings' threshold (successful firms) than it is to other firms (unsuccessful firms).

To address this issue, we define DS, a dummy variable taking the value 1 ifpre-managed earnings and after-managed earnings are higher than prior year's earningsand 0 otherwise. We compute pre-managed earnings as net income minus discretionaryaccruals as estimated by Eq. (1) after we have multiplied them with their lagged totalassets. After-managed earnings are simply the net income reported by a firm. We theninteract DS with ETRCFO and re-estimate our equations.

Table 9 summarizes these findings. The reported results indicate that regardless of thesign of discretionary accruals, successful firms exhibit a more pronounced negativerelation with discretionary accruals; DS is significantly negative and takes a value of−0.009 and −0.006 in the case of positive and negative discretionary accruals, respectively.This relationship is expected because Greek firms are well known for striving for smoothedearnings (Leuz et al., 2003). These efforts are likely induced by the unwillingness to increasedividends (Kasanen et al., 1996) and the intention to create cookie-jar reserves (Levitt, 1998).Interestingly, the incremental effect of IFRS adoption appears insignificant, which suggeststhat this negative relation continues to hold in the post-IFRS period.

Consistent with expectations, regardless of the sign of discretionary accruals, thesignificance of the tax pressure varies across successful and unsuccessful firms in thepre-IFRS period; the successful firms exhibit an incremental reaction. In particular,DS ∗ ETRCFO takes values of −0.061 and −0.049 for positive and negative discretionaryaccruals, respectively. This finding indicates that tax-induced incentives to restrictaggressive financial reporting or exacerbate downward earnings management are amplifiedwhen a firm has passed the previous year earnings' threshold. This finding further supportsprior inferences that tax pressure constitutes an economic determinant of financial earningsmanagement in the pre-IFRS period and that the previously documented effect is not amere spurious relationship. However, in the post-IFRS period, the incremental coefficient(DIFRS ∗ DS ∗ ETRCFO) loads with a positive sign, thereby confirming priorconclusions that the effect of tax pressure weakens after IFRS adoption.

Notes to Table 9:Variable definitions: DACC+ is positive discretionary accruals; DACC− is negative discretionary accruals; DS is adummy variable taking the value 1 in the case of a firm whose current pre-managed and reported earnings arehigher than the previous year's reported earnings and 0 otherwise; ETRCFO is current tax scaled by cash flowfrom operations; SIZE is the natural log of total assets; LEV is total liabilities minus cash, scaled by total assets;CURRENT is current assets scaled by current liabilities; CFO is cash flow from operations scaled by lagged totalassets; AUD is a dummy variable taking the value 1 if a firm is audited by a Big 4 firm and 0 otherwise; IPO is adummy variable taking the value 1 if a firm goes public in year t and 0 otherwise; and DIFRS is a dummy variabletaking the value 1 in the post-IFRS period and 0 otherwise.⁎⁎⁎ Represents significance at the 1% level.⁎⁎ Represents significance at the 5% level.⁎ Represents significance at the 10% level.


As a final test, we investigate whether the number of successful firms with high taxpressure has increased following IFRS adoption.20 In the pre-IFRS period, high taxpressure is likely to have deterred more firms from managing income upward to meetearnings targets. If the importance of tax pressure declined after IFRS adoption, the rate offirms with high tax pressure that pass prior year earnings' threshold may have increased.

To address this point, we restrict our sample to successful firms and define HighETRCFO,a dummy variable taking the value 1 if a successful firm's ETRCFO is above the meanETRCFO in a year. Table 10 illustrates a two-by-two analysis between DIFRS and thefrequency of HighETRCFO firms. The reported results indicate that 43.20% of successfulfirms were above the annual mean ETRCFO in the pre-IFRS period. This percentageincreases to 57.19% after IFRS adoption. A chi-squared test reported at the bottom of the tablesuggests that the relationship between DIFRS and the frequency of HighETRCFO successfulfirms is statistically significant at the 5% level, thereby supporting our earlier inferences thatthe importance of tax-pressure declines after IFRS adoption.

8. Summary and conclusions

In this study, we examine the effects of changes in book–tax conformity on tax-inducedincentives for financial earnings management.We use a sample of Greek firms reported undera high book–tax conformity regime prior to IFRS adoption. IFRS enactment reduced book–tax conformity and released financial income from direct tax implications. Consistent withexpectations, we find that in the pre-IFRS period, tax pressure exhibits a significantlynegative relationship with positive and negative discretionary accruals. In the post-IFRSperiod, we find that this relationship dissipates. Furthermore, we also document anincremental reaction to tax pressure for firms that succeed in meeting the previous year'searnings when they report under high book–tax conformity. Therefore, we infer that highbook–tax conformity amplifies tax-induced incentives to restrict (exacerbate) upward(downward) financial earnings management for firms with higher tax pressure. Furthermore,we conclude that IFRS adoption profoundly affects tax-induced incentives for financialearnings management through the enforced separation of financial and tax reporting; a greaterseparation of the two systems attenuates tax-induced incentives and releases financialreporting aggressiveness. We note that an alternative explanation for these results may be areduction in statutory tax rates occurring in the post-IFRS period. However, additionalanalysis of a matched sample of private firms supports our initial inferences.

However, our analysis is limited to the research field of earnings management. We believethat our findings have further implications for other aspects of accounting earnings (e.g.,earnings informativeness, conditional conservatism, information asymmetry) and areworthwhile subjects for future research. It is likely that the dissipation of tax-inducedincentives to manipulate financial income would benefit these market attributes. For instance,empirical evidence provided by Hanlon et al. (2006), and Hanlon et al. (2008) suggests thathigh book–tax conformity has a negative impact on earnings informativeness. However,Karampinis and Hevas (2011) report that mandatory IFRS adoption in Greece has notimproved earnings quality (measured as the value relevance and the conditional conservatism

20 We would like to thank an anonymous referee for raising this point.

Table 10Two-by-two analysis of successful firms with high ETRCFO versus successful firms with low ETRCFO.

HighETRCFO = 0 HighETRCFO = 1 Total

DIFRS = 0 117 89 20656.80% 43.20% 100%

DIFRS = 1 110 147 25742.80% 57.20% 100%

Total 227 236

Test of the null hypothesis that there is no relationship between IFRS adoption and the frequency of successfulfirms with high ETRCFO

χ2 = 8.961 p-Value = 0.01

Notes:Variable definitions: DIFRS is a dummy variable taking the value 1 in the post-IFRS period and 0 otherwise; andHighETRCFO is a dummy variable taking the value 1 if a successful firm's ETRCFO is above the annual meanETRCFO and 0 otherwise.


of accounting earnings) in a material way. This finding implies that relaxations in book–taxconformity may be beneficial only in settings where the other institutional factors favorablyinduce managers to provide high-quality accounting information. To the extent that thiscondition does not hold in Greece, managers may simply replace tax-induced motivations forfinancial income management with other opportunistic incentives.21

Despite this limitation, we believe our study makes a significant contribution to theexisting literature. In particular, it adds to the still-scarce studies on book–tax conformity,because it illuminates the reaction of managerial opportunism to changes in taximplications of financial income. In addition, it provides timely insights into potentialimpacts of IFRS adoption in code-law countries with strong tax-orientation.22 Overall, ourfindings suggest that the reduction in book–tax conformity imbedded by IFRS adoptionlimits the influence of an important determinant of managerial opportunism with directramifications for financial earnings.

References

Ali, A., & Hwang, L. (2000). Country-specific factors related to financial reporting and the value relevance ofaccounting data. Journal of Accounting Research, 38(1), 1–21.

Arnedo, L., Lizarraga, F., & S nchez, S. (2007). Does public/private status affect the level of earningsmanagement in code-law contexts outside the United States? A study based on the Spanish case. TheInternational Journal of Accounting, 42(3), 305–328.

Atwood, T., Drake, M., & Myers, L. (2010). Book–tax conformity, earnings persistence and the associationbetween earnings and future cash flows. Journal of Accounting and Economics, 50(1), 111–125.

Atwood, T. J., Drake, M. S., Myers, J. N., & Myers, L. A. (2011). Do earnings reported under IFRS tell us moreabout future earnings and cash flows? Journal of Accounting and Public Policy, 30(2), 103–121.

21 See for instance Jeanjean and Stolowy (2008) who report that IFRS adoption increased earnings managementin France.22 See Schanz and Schanz (2010) for a pertinent discussion.


Ball, R., & Shivakumar, L. (2005). Earnings quality in UK private firms: Comparative loss recognition timeliness.Journal of Accounting and Economics, 39(1), 83–128.

Ballas, A., & Hevas, D. (2000). The evolution of [tax] accounting in Greece: Working paper presented in the 23rdAnnual Congress of the European Accounting Association in Munich, Germany.

Bartov, E., Gul, F., & Tsui, J. (2000). Discretionary-accruals models and audit qualifications. Journal ofAccounting and Economics, 30(3), 421–452.

Becker, C., DeFond, M., Jiambalvo, J., & Subramanyam, K. (1998). The effect of audit quality on earningsmanagement. Contemporary Accounting Research, 15, 1–24.

Boynton, C., Dobbins, P., & Plesko, G. (1992). Earnings management and the corporate alternative minimum tax.Journal of Accounting Research, 30, 131–153.

Burgstahler, D., & Dichev, I. (1997). Earnings management to avoid earnings decreases and losses. Journal ofAccounting and Economics, 24(1), 99–126.

Burgstahler, D., Hail, L., & Leuz, C. (2006). The importance of reporting incentives: Earnings management inEuropean private and public firms. The Accounting Review, 81, 983–1016.

Bushman, R., & Piotroski, J. (2006). Financial reporting incentives for conservative accounting: The influence oflegal and political institutions. Journal of Accounting and Economics, 42(1–2), 107–148.

Butler, M., Leone, A., & Willenborg, M. (2004). An empirical analysis of auditor reporting and its associationwith abnormal accruals. Journal of Accounting and Economics, 37(2), 139–165.

Byard, D., Li, Y., & Yu, Y. (2011). The effect of mandatory IFRS adoption on financial analysts' informationenvironment. Journal of Accounting Research, 49(1), 69–96.

Caramanis, C., & Lennox, C. (2008). Audit effort and earnings management. Journal of Accounting andEconomics, 45(1), 116–138.

Chang, J., Khanna, T., & Palepu, K. (2000). Analyst activity around the world: Working paper. University ofPennsylvania.

Chen, S., Chen, X., Cheng, Q., & Shevlin, T. (2010). Are family firms more tax aggressive than non-family firms?Journal of Financial Economics, 95(1), 41–61.

Cook, K. A., Huston, G. R., & Omer, T. C. (2008). Earnings management through effective tax rates: The effectsof tax-planning investment and the Sarbanes–Oxley Act of 2002. Contemporary Accounting Research, 25(2),447–471.

Coppens, L., & Peek, E. (2005). An analysis of earnings management by European private firms. Journal ofInternational Accounting, Auditing and Taxation, 14(1), 1–17.

Daske, H., Hail, L., Leuz, C., & Verdi, R. (2008). Mandatory IFRS reporting around the world: Early evidence onthe economic consequences. Journal of Accounting Research, 46(5), 1081–1142.

Dechow, P. (1994). Accounting earnings and cash flows as measures of firm performance: The role of accountingaccruals. Journal of Accounting and Economics, 18(1), 3–42.

Dechow, P., Sloan, R., & Sweeney, A. (1995). Detecting earnings management. The Accounting Review, 70(2),193–225.

DeFond, M., & Jiambalvo, J. (1994). Debt covenant violation and manipulation of accruals. Journal ofAccounting and Economics, 17(1–2), 145–176.

Desai, M. A. (2005). The degradation of reported corporate profits. The Journal of Economic Perspectives, 19(4),171–192.

Dhaliwal, D., & Wang, S. (1992). The effect of book income adjustment in the 1986 alternative minimum tax oncorporate financial reporting. Journal of Accounting and Economics, 15(1), 7–26.

Ding, Y., Hope, O., Jeanjean, T., & Stolowy, H. (2007). Differences between domestic accounting standards andIAS: Measurement, determinants and implications. Journal of Accounting and Public Policy, 26(1), 1–38.

Dyreng, S. D., Hanlon, M., & Maydew, E. L. (2008). Long-run corporate tax avoidance. The Accounting Review,83(1), 61–82.

Eberhartinger, E. L. E. (1999). The impact of tax rules on financial reporting in Germany, France, and the UK. TheInternational Journal of Accounting, 34(1), 93–119.

European Commission (1998). Implementation of the Fourth Directive in EU member states.Fan, J., & Wong, T. (2002). Corporate ownership structure and the informativeness of accounting earnings in East

Asia. Journal of Accounting and Economics, 33(3), 401–425.Francis, J., & Wang, D. (2008). The joint effect of investor protection and Big 4 audits on earnings quality around

the world. Contemporary Accounting Research, 25(1), 157–191.


Frank, M., Lynch, L., & Rego, S. (2009). Tax reporting aggressiveness and its relation to aggressive financialreporting. The Accounting Review, 84(2), 467–496.

Fullerton, D. (1984). Which effective tax rate? National Tax Journal, 37(March), 23–42.Ghicas, D., Papadaki, A., Siougle, G., & Sougiannis, T. (2008). The relevance of quantifiable audit qualifications

in the valuation of IPOs. Review of Accounting Studies, 13(4), 512–550.Graham, J. R. (1996a). Debt and the marginal tax rate. Journal of Financial Economics, 41(1), 41–73.Graham, J. R. (1996b). Proxies for the corporate marginal tax rate. Journal of Financial Economics, 42(2),

187–221.Grigorakos, T. (2006). Journal entries for income distribution accomplished by listed firms that keep their records

in accordance with GAS but report under IFRS (in Greek). Logistis, 1, 29–37.Guenther, D. (1994). Earnings management in response to corporate tax rate changes: Evidence from the 1986

Tax Reform Act. The Accounting Review, 69(1), 230–243.Guenther, D., Maydew, E., & Nutter, S. (1997). Financial reporting, tax costs, and book–tax conformity. Journal

of Accounting and Economics, 23(3), 225–248.Guenther, D., & Young, D. (2000). The association between financial accounting measures and real economic

activity: A multinational study. Journal of Accounting and Economics, 29(1), 53–72.Gupta, S., & Newberry, K. (1997). Determinants of the variability in corporate effective tax rates: Evidence from

longitudinal data. Journal of Accounting and Public Policy, 16(1), 1–34.Hanlon, M. (2003). What can we infer about a firm's taxable income from its financial statements? National Tax

Journal, 56(4), 831–864.Hanlon, M. (2005). The persistence and pricing of earnings, accruals, and cash flows when firms have large book–

tax differences. The Accounting Review, 80(1), 137–166.Hanlon, M., & Heitzman, S. (2010). A review of tax research. Journal of Accounting and Economics, 50(2–3),

127–178.Hanlon, M., Laplante, S., & Shevlin, T. (2005). Evidence for the possible information loss of conforming book

income and taxable income. Journal of Law and Economics, 48(2), 407–442.Hanlon, M., Maydew, E., & Shevlin, T. (2006). Book–tax conformity and the informativeness of earnings:

Working paper. University of Michigan.Hanlon, M., Maydew, E., & Shevlin, T. (2008). An unintended consequence of book–tax conformity: A loss of

earnings informativeness. Journal of Accounting and Economics, 46(2–3), 294–311.Hanlon, M., & Shevlin, T. (2005). Book–tax conformity for corporate income: An introduction to the issues. Tax

Policy and the Economy, 101–134.Hevas, D. (1995). Financial accounting, tax accounting and the fourth directive: A critical analysis of Greece's

response to accounting harmonisation: Working paper presented in workshop on accounting in Geneva,Switzerland.

Hung, M. (2000). Accounting standards and value relevance of financial statements: An international analysis.Journal of Accounting and Economics, 30(3), 401–420.

Hung, M., & Subramanyam, K. (2007). Financial statement effects of adopting international accounting standards:The case of Germany. Review of Accounting Studies, 12(4), 623–657.

Jeanjean, T., & Stolowy, H. (2008). Do accounting standards matter? An exploratory analysis of earningsmanagement before and after IFRS adoption. Journal of Accounting and Public Policy, 27(6), 480–494.

Jones, J. (1991). Earnings management during import relief investigations. Journal of Accounting Research,29(2), 193–228.

Jones, K., Krishnan, G., & Melendrez, K. (2008). Do discretionary accruals models detect actual cases offraudulent and restated earnings? An empirical evaluation. Contemporary Accounting Research, 25,499–531.

Joos, P., & Lang, M. (1994). The effects of accounting diversity: Evidence from the European Union. Journal ofAccounting Research, 32, 141–168.

Karampinis, N., & Hevas, D. (2011). Mandating IFRS in an unfavorable environment: The Greek experience. TheInternational Journal of Accounting, 46(3), 304–332.

Kasanen, E., Kinnunen, J., & Niskanen, J. (1996). Dividend-based earnings management: Empirical evidencefrom Finland. Journal of Accounting and Economics, 22(1–3), 283–312.

Kothari, S., Leone, A., & Wasley, C. (2005). Performance matched discretionary accrual measures. Journal ofAccounting and Economics, 39(1), 163–197.


La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (1997). Legal determinants of external finance.Journal of Finance, 52(3), 1131–1150.

Landsman, W. R., Maydew, E. L., & Thornock, J. R. (2011). The information content of annual earningsannouncements and mandatory adoption of IFRS. Journal of Accounting and Economics, 53(1), 34–54.

Lara, J., Osma, B., & Mora, A. (2005). The effect of earnings management on the asymmetric timeliness ofearnings. Journal of Business Finance and Accounting, 32(3–4), 691–726.

Leuz, C., Nanda, D., & Wysocki, P. (2003). Earnings management and investor protection: An internationalcomparison. Journal of Financial Economics, 69(3), 505–527.

Levitt, A. (1998). The numbers game: Remarks delivered at the NYU Center for Law and Business, New York.Li, S. (2010). Does mandatory adoption of international financial reporting standards in the European Union

reduce the cost of equity capital? The Accounting Review, 85, 607.Lin, K. Z. (2006). The impact of tax holidays on earnings management: An empirical study of corporate

reporting behavior in a developing-economy framework. The International Journal of Accounting, 41(2),163–175.

Lopez, T., Regier, P., & Lee, T. (1998). Identifying tax-induced earnings management around TRA 86 as afunction of prior tax-aggressive behavior. Journal of American Taxation Association, 20(2), 37–56.

Maydew, E. (1997). Tax-induced earnings management by firms with net operating losses. Journal of AccountingResearch, 35(1), 83–96.

McNichols, M. (2000). Research design issues in earnings management studies. Journal of Accounting and PublicPolicy, 19(4–5), 313–345.

Nobes, C. (2001). GAAP 2001: A survey of national accounting rules benchmarked against internationalaccounting standards: International forum on accountancy development.

Nobes, C., & Parker, R. (2008). Comparative international accounting. Financial Times/Prentice Hall.Othman, H. B., & Zeghal, D. (2006). A study of earnings-management motives in the Anglo-American and

Euro-Continental accounting models: The Canadian and French cases. The International Journal ofAccounting, 41(4), 406–435.

Papas, A. (1992). Accounting in Greece. London: Routledge.Peasnell, K., Pope, P., & Young, S. (2000). Detecting earnings management using cross-sectional abnormal

accrual models. Accounting and Business Research, 30(4), 313–326.Petersen, M. (2009). Estimating standard errors in finance panel data sets: Comparing approaches. Review of

Financial Studies, 22(1), 435–480.Phillips, J., Pincus, M., & Rego, S. (2003). Earnings management: New evidence based on deferred tax expense.

The Accounting Review, 78(2), 491–521.Phillips, J., Pincus, M., Rego, S., & Wan, H. (2004). Decomposing changes in deferred tax assets and liabilities to

isolate earnings management activities. Journal of American Taxation Association, 26, 43–66.Porcano, T. M., & Tran, A. V. (1998). Relationship of tax and financial accounting rules in Anglo-Saxon

countries. The International Journal of Accounting, 33(4), 433–454.Rajan, R., & Zingales, L. (2003). Banks and markets: The changing character of European finance. Centre for

Economic Policy Research.Roubi, R., & Richardson, W. (1998). Managing discretionary accruals in response to reductions in corporate tax

rates in Canada, Malaysia and Singapore. The International Journal of Accounting, 33(4), 455–467.Schanz, D., & Schanz, S. (2010). Finding a new corporate tax base after the abolishment of the one-book system

in EU member states. The European Accounting Review, 19(2), 311–341.Scholes, M., Wilson, G., & Wolfson, M. (1992). Firms' responses to anticipated reductions in tax rates: The Tax

Reform Act of 1986. Journal of Accounting Research, 30, 161–185.Shevlin, T. (1990). Estimating corporate marginal tax rates with asymmetric tax treatment of gains and losses.

Journal of American Taxation Association, 12(1), 51–67.Stickney, C., & McGee, V. (1982). Effective corporate tax rates the effect of size, capital intensity, leverage, and

other factors. Journal of Accounting and Public Policy, 1(2), 125–152.Subramanyam, K. (1996). The pricing of discretionary accruals. Journal of Accounting and Economics, 22(1–3),

249–281.Teoh, S., Wong, T., & Rao, G. (1998). Are accruals during initial public offerings opportunistic? Review of

Accounting Studies, 3(1), 175–208.Thanopoulou, E. (2001). Introduction in financial accounting, volume B' (in Greek) (G. Benou). (Editions).


Van Tendeloo, B., & Vanstraelen, A. (2005). Earnings management under German GAAP versus IFRS. TheEuropean Accounting Review, 14(1), 155–180.

Vander Bauwhede, H., Willekens, M., & Gaeremynck, A. (2003). Audit firm size, public ownership, and firms'discretionary accruals management. The International Journal of Accounting, 38(1), 1–22.

Wang, S. (1994). The relationship between financial reporting practices and the 1986 alternative minimum tax.The Accounting Review, 69(3), 495–506.

Wang, D. (2006). Founding family ownership and earnings quality. Journal of Accounting Research, 44(3),619–656.

Watts, R. (2003). Conservatism in accounting part I: Explanations and implications. Accounting Horizons, 17(3),207–222.

Watts, R., & Zimmerman, J. (1986). Positive accounting theory. New York: Prentice Hall, International Editions.Young, S. (1999). Systematic measurement error in the estimation of discretionary accruals: An evaluation of

alternative modelling procedures. Journal of Business Finance and Accounting, 26(7&8), 833–862.Zimmerman, J. (1983). Taxes and firm size. Journal of Accounting and Economics, 5, 119–149.

effects of ifrs adoption on tax-induced incentives for financial earnings management: evidence from...

Documents