198Oxford Review of Economic Policy vol. 21 no. 2 2005

© The Author (2005). Published by Oxford University Press. All rights reserved.

A THEORY OF CORPORATESCANDALS: WHY THE USAAND EUROPE DIFFER

OXFORD REVIEW OF ECONOMIC POLICY, VOL. 21, NO. 2DOI: 10.1093/oxrep/gri012

JOHN C. COFFEE, JRColumbia University Law School

A wave of financial irregularity in the USA in 2001–2 culminated in the Sarbanes–Oxley Act. A worldwide stock-market bubble burst over this same period, with the actual market decline being proportionately more severe inEurope. Yet, no corresponding wave of financial scandals involving a similar level of companies occurred inEurope. Given the higher level of public and private enforcement in the USA for securities fraud, this contrast seemsperplexing. This paper submits that different kinds of scandals characterize different systems of corporate govern-ance. In particular, dispersed ownership systems of governance are prone to the forms of earnings management thaterupted in the USA, but concentrated ownership systems are much less vulnerable. Instead, the characteristicscandal in such systems is the appropriation of private benefits of control. This paper suggests that this differencein the likely source of, and motive for, financial misconduct has implications both for the utility of gatekeepers asreputational intermediaries and for design of legal controls to protect public shareholders. The difficulty inachieving auditor independence in a corporation with a controlling shareholder may also imply that minorityshareholders in concentrated ownership economies should directly select their own gatekeepers.

I. INTRODUCTION

Corporate scandals, particularly when they occur inconcentrated outbursts, raise serious issues thatscholars have too long ignored. Two issues standout. First, why do different types of scandals occurin different economies? Second, why does a waveof scandals occur in one economy, but not in an-other, even though both economies are closelyinterconnected in the same global economy and

subject to the same macroeconomic conditions?This brief essay seeks to relate answers to bothquestions to the structure of share ownership.

Conventional wisdom explains a sudden concentra-tion of corporate financial scandals as the conse-quence of a stock-market bubble. When the bubblebursts, scandals follow, and, eventually, new regu-lation.1 Historically, this has been true at least sincethe South Seas Bubble, and this hypothesis works

1 For a pre-Sarbanes–Oxley review of the last 300 years of this pattern, see Banner (1997).

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

reasonably well to explain the turn-of-the-millen-nium experience in the USA and Europe. World-wide, a stock-market bubble did burst in 2000, andin percentage terms the decline was greater in manyEuropean countries than in the United States.2 Butin Europe, this sudden market decline was notassociated with the same pervasive accounting andfinancial irregularity that shook the US economyand produced the Sarbanes–Oxley Act in 2002.Indeed, financial statement restatements are rare inEurope.3 In contrast, the USA witnessed an accel-erating crescendo of financial statement restate-ments that began in the late 1990s. The UnitedStates General Accounting Office (GAO) has foundthat over 10 per cent of all listed companies in theUnited States announced at least one financialstatement restatement between 1997 and 2002(GAO, 2002, p. 4). Later studies have placed thenumber even higher (Huron Consulting Group(2003a) discussed at the beginning of section II).Because a financial statement restatement is aserious event in the United States that, depending onits magnitude, often results in a private class action,a Securities and Exchange Commission (SEC) en-forcement proceeding, a major stock price drop,and/or a management shake-up, one suspects thatthese announced restatements were but the tip ofthe proverbial iceberg, with many more companiesnegotiating changes in their accounting practiceswith their outside auditors that averted a formalrestatement.

While Europe also had financial scandals over thissame period (with the Parmalat scandal being themost notorious4), most were characteristically dif-ferent from the US style of earnings manipulationscandal (of which Enron and WorldCom were theiconic examples). Only European firms cross-listedin the United States seem to have encounteredsimilar crises of earnings management (see section

II para. 5). What explains this difference and thedifference in frequency? This essay advances asimple, almost self-evident thesis: differences in thestructure of share ownership account for differ-ences in corporate scandals, both in terms of thenature of the fraud, the identity of the perpetrators,and the seeming disparity in the number of scandalsat any given time. In dispersed ownership systems,corporate managers tend to be the rogues of thestory, while in concentrated ownership systems, it iscontrolling shareholders who play the correspond-ing role. Although this point may seem obvious, itscorollary is less so: the modus operandi of fraud isalso characteristically different. Corporate manag-ers tend to engage in earnings manipulation, whilecontrolling shareholders tend to exploit the privatebenefits of control. Finally, and most importantly,given these differences, the role of gatekeepers inthese two systems must necessarily also be differ-ent.5 While gatekeepers failed both at Enron andParmalat, they failed in characteristically differentways. In turn, different reforms may be justified,and the panoply of reforms adopted in the UnitedStates, culminating in the Sarbanes–Oxley Act of2002, may not be the appropriate remedy in Europe.

Section II reviews the recent American scandals toidentify common denominators and the underlyingmotivation that caused the sudden eruption of finan-cial statement restatements. Section III turns to theevidence on private benefits of control in concen-trated ownership systems. Patterns also emergehere in terms of the maturity of the capital market.Section IV advances some tentative conclusionsabout the differences in monitoring structures thatare appropriate under different ownership regimes.At the outset, however, it should be underscoredthat companies with dispersed ownership and com-panies with concentrated ownership co-exist in allmajor jurisdictions.6 Thus, we are talking about

2 See Holmstrom and Kaplan (2003), who show that from 2001 through 31 December 2002, the US stock-market returns were–32 per cent, while France was –45 per cent, and Germany –53 per cent.

3 Although they have been rare in the past, FitchRatings, the credit ratings agency, predicts that they will become common inEurope in 2005, as thousands of European companies switch from local accounting standards to International Financial ReportingStandards, which are more demanding. See FitchRatings (2005).

4 For a detailed review of the Parmalat scandal, see Melis (2004).5 The term ‘gatekeeper’ will not be elaborately defined for the purposes of this short essay, but means a reputational intermediary

who pledges its considerable reputational capital to give credibility to its statements or forecasts. Auditors, securities analysts,and credit ratings agencies are the most obvious examples. See Coffee (2004a).

6 This has been demonstrated at length. See La Porta et al. (1999).

199

J. C. Coffee, Jr

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

200

OXFORD REVIEW OF ECONOMIC POLICY, VOL. 21, NO. 2

tendencies and not any iron law that admits of noexception. As much scholarship has demonstrated,the corporate universe divides, more or less, into twobasically alternative systems of corporate govern-ance:

(i) a dispersed ownership system, characterizedby strong securities markets, rigorous disclo-sure standards, high share turnover, and highmarket transparency, in which the market forcorporate constitutes the ultimate disciplinarymechanism; and

(ii) a concentrated ownership system, charac-terized by controlling blockholders, weaker se-curities markets, high private benefits of con-trol, and lower disclosure and market transpar-ency standards, but with a possiblysubstitutionary role played by large banks andnon-controlling blockholders.7

This essay advances only the idea that the role ofgatekeepers also differs across these legal regimesand that gatekeepers, including even the auditor,arguably play a more central and critical role in thedispersed ownership system.

II. FRAUD IN DISPERSEDOWNERSHIP SYSTEMS

While studies differ, all show a rapid acceleration infinancial statement restatements in the United Statesduring the 1990s. The earliest of these studies findsthat the number of earnings restatements by publiclyheld US corporations averaged roughly 49 per yearfrom 1990 to 1997, increased to 91 in 1998, and thensoared to 150 and 156 in 1999 and 2000, respectively(see Moriarty and Livingston, 2001, pp. 53–4). Alater study by the United States GAO shows aneven more dramatic acceleration, as set forth inFigure 1 (GAO, 2002).

Even this study understated the severity of thissudden spike in accounting irregularity. Becausecompanies do not uniformly report a restatement inthe same fashion, the GAO was not able to catch allrestatements in its study. A more recent, fuller studyin 2003 by Huron Consulting Group8 shows thefollowing results: in 1990, there were 33 earningsrestatements; in 1995, there were 50 (Huron Con-sulting Group, 2003a); then, the rate truly acceler-ated to 216 in 1999; to 233 in 2000; to 270 in 2001;and then in 2002, the number peaked at 330 (ten

7 For an overview of these rival systems, see Coffee (2001).8 See Huron Consulting Group (2003b, p. 4). The number of restatements fell (slightly) to 323 in 2003 (Huron Consulting Group,

2003b). Not all these restatements involved overstatements of earnings (and some involved understatements). Still, the rising rateof restatements seems a good proxy for financial irregularity.

Figure 1Total number of Restatement Announcements Identified, 1997–2002

92 102

174201

225

125

0

50

100

150

200

250

300

1997 1998 1999 2000 2001 2002 (as of30 June)

250 (projected for year end)

(act

ual)

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

201

J. C. Coffee, Jr

times the 1990 level). On this basis, roughly one ineight listed companies restated over this period. Anupdate this year by Huron Consulting shows that thenumber of restatements fell to 323 in 2003 and thenrose again to 414 in 2004.9 In any event, even if theexact number of restatements is disputed, theoverall pattern of a hyperbolic rate of increasearound the turn of the millennium persists acrossall studies.

Nor were these restatements merely technical ad-justments. Although some actually increased earn-ings, the GAO study found that the typical restatingfirm lost an average 10 per cent of its marketcapitalization over a 3-day trading period surround-ing the date of the announcement (GAO, 2002, p. 5).All told, the GAO estimated the total market losses(unadjusted for other market movements) at $100billion for restating firms in its incomplete sample for1997–2002 (GAO, 2002, p. 34).

Other studies have reached similar results. Studyinga comprehensive sample of firms that restatedannual earnings from 1971 to 2000, Richardson etal. (2002, p. 4) reported a negative market reactionto the announcement of the restatement of 11 percent over a 3-day window surrounding the an-nouncement. Moreover, using a wider window thatmeasured firm value over a period from 120 daysprior to the announcement to 120 days after it, theyfound that restating ‘firms lose on average 25 percent of market value over the period examined andthis is concentrated in a narrow window surroundingthe announcement of the restatement’ (Richardsonet al., 2002, p. 16). Twenty-five per cent of marketvalue represents an extraordinary market penalty. Itshows the market not simply to have been surprised,but to have taken the restatement as a signal offraud. For example, in the cases of Cendant,MicroStrategy, and Sunbeam, three major US cor-porate scandals in the late 1990s, they found that‘these three firms lost more than $23 billion in theweek surrounding their respective restatement an-nouncements’ (Richardson et al., 2002).

The intensity of the market’s negative reaction to anearnings restatement varied with the cause of the

restatement, with the most negative reactions beingassociated with restatements involving revenue rec-ognition issues (see Anderson and Yohn, 2002).One study, examining just the period from 1997 to1999, found that firms in which revenue recognitionissues caused the restatement experienced a wellabove average, market-adjusted loss of –13.38 percent over a window period beginning 3 days beforethe announcement and continuing until 3 days afterit.10 Revenue recognition errors essentially revealedmanagement not just to have been mistaken, but tohave cheated, and the market reacted accordingly.Yet, despite the market’s fear of such practices,revenue recognition errors became the dominantcause of restatements in the period from 1997 to2002. The GAO Report found that revenue recog-nition issues accounted for almost 38 per cent of therestatements it identified over that period (GAO,2002, p. 28), and the Huron Consulting Group studyalso found it to be the leading accounting issueunderlying an earnings restatement between 1999and 2003 (Huron Consulting Group, 2003a, p. 4).

The prevalence of revenue recognition problems,even in the face of the market’s sensitivity to them,shows a significant change in managerial behaviourin the United States. During earlier periods, USmanagements famously employed ‘rainy day re-serves’ to hold back the recognition of income thatwas in excess of the market’s expectation in orderto defer its recognition until some later quarter whenthere had been a shortfall in expected earnings. Ineffect, managers engaged in income-smoothing,rolling the peaks in one period over into the valley ofthe next period. This traditional form of earningsmanagement was intended to mask the volatility ofearnings and reassure investors who might havebeen alarmed by rapid fluctuations in earnings. Incontrast, managers in the late 1990s appear to havecharacteristically ‘stolen’ earnings from future pe-riods in order to create an earnings spike thatpotentially could not be sustained. Why? Although ithad long been known that restating firms weretypically firms with high market expectations forfuture growth, the pressure on these firms to showa high rate of earnings growth appears to haveincreased during the 1990s.

9 See Huron Consulting Group (2005). If one wishes to focus only on restatements of the annual audited financial statements,excluding restatements of quarterly earnings, the numbers were: 2000, 98; 2001, 140; 2002, 183; 2003, 206; 2004, 253.

10 Anderson and Yohn (2002, p. 13). This loss was measured in terms of cumulative abnormal returns (CAR). Where the causeof the restatement was reported as ‘fraud’, the CAR rose to –19 per cent, but there were only a handful of such cases.

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

202

OXFORD REVIEW OF ECONOMIC POLICY, VOL. 21, NO. 2

What, in turn, caused this increased pressure? To aconsiderable extent, it appears to have been self-induced—that is, the product of increasingly opti-mistic predictions by managements to financial ana-lysts as to future earnings. But this answer justtranslates the prior question into a different format:why did managements become more optimisticabout earnings growth over this period? Here, oneexplanation does distinguish the USA from Europe,and it has increasingly been viewed as the bestexplanation for the sudden spike in financial irregu-larity in the USA.11 Put simply, executive compen-sation abruptly shifted in the United States duringthe 1990s, moving from a cash-based system to anequity-based system. More importantly, this shiftwas not accompanied by any compensating changein corporate governance to control the predictablyperverse incentives that reliance on stock optionscan create.

One measure of the suddenness of this shift is thechange over the decade in the median compensationof a chief executive officer (CEO) of an S&P 500industrial company. As of 1990, the median suchCEO made $1.25m with 92 per cent of that amountpaid in cash and 8 per cent in equity (see Hall, 2003).But during the 1990s, both the scale and compositionof executive compensation changed. By 2001, the

median CEO of an S&P industrial company wasearning over $6m, of which 66 per cent was in equity(Hall, 2003). Figure 2 shows the swiftness of thistransition (Hall, 2003).

To illustrate the impact of this change, assume aCEO holds options on 2m shares of his company’sstock and that the company is trading at a price toearnings ratio of 30 to 1 (both reasonable assump-tions for this era). On this basis, if the CEO cancause the ‘premature’ recognition of revenues thatresult in an increase in annual earnings by simply $1per share, the CEO has caused a $30 price increasethat should make him $60m richer. Not a smallincentive!

Obviously, when one pays the CEO with stockoptions, one creates incentives for short-term finan-cial manipulation and accounting gamesmanship.Financial economists have found a strong statisticalcorrelation between higher levels of equity compen-sation and both earnings management and financialrestatements. One recent study by Efendi et al.(2004) utilized a control group methodology andconstructed two groups of companies, each com-posed of 100 listed public companies.12 The firstgroup’s members had restated their financial state-ments in 2001 or 2002, while the control group was

11 For a fuller account of these various explanations, see Coffee (2004b).12 Efendi et al. (2004). For an earlier similar study, see Johnson et al. (2003).

Figure 2CEO Compensation at S&P 500 Industrial Companies, 1980–2001

Source: Brian J. Hall, Harvard Business School.

0

1

2

3

4

5

6

7

8

1980 1983 1986 1989 1992 1995 1998 2001

Med

ian

CEO

pay

($m

)

Cash Equity

32% 37

% 40% 43

%

49%

54%

58%

60%

63%

66%

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

203

J. C. Coffee, Jr

composed of otherwise similar firms that had notrestated. What characteristic most distinguished thetwo groups? The leading factor that proved most toinfluence the likelihood of a restatement was thepresence of a substantial amount of ‘in the money’stock options in the hands of the firm’s CEO. TheCEOs of the firms in the restating group held onaverage ‘in the money’ options of $30.9m, whileCEOs in the non-restating control group averagedonly $2.3m—a nearly 14 to 1 difference (Efendi etal., 2004, p. 2). Further, if a CEO held optionsequalling or exceeding 20 times his or her annualsalary (and this was the 80th percentile in theirstudy—meaning that a substantial number of CEOsdid exceed this level), the likelihood of a restatementincreased by 55 per cent.

Other studies have reached similar results. Denis etal. (2005) find a significant positive relationshipbetween a firm’s use of option-based compensationand securities fraud allegations being levelled againstthe firm.13 Further, they find in their study of 358companies charged with fraud between 1993 and2002 that the likelihood of a fraud charge is posi-tively related to ‘option intensity’—i.e. the greaterthe amount of the options, the higher the likelihood(Denis et al., 2005, p. 4). Similarly, Cheng andWarfield (2004) have documented that corporatemanagers with high equity incentives sell moreshares in subsequent periods, are more likely toreport earnings that just meet or exceed analysts’forecasts, and more frequently engage in otherforms of earnings management. As stock optionsincrease the managers’ equity ownership, they alsoincrease their need to diversify the high risk associ-ated with such ownership, and this produces bothmore efforts to inflate earnings to prevent a stockprice decline and increased sales by managers inadvance of any earnings decline. In short, there is a‘dark side’ to option-based compensation for seniorexecutives: absent special controls, more optionsmean more fraud.

At this point, the contrast between managerialincentives in the USA and Europe comes into

clearer focus. These differences involve both thescale of compensation and its composition. In 2004,CEO compensation as a multiple of average em-ployee compensation was estimated to be 531:1 inthe USA, but only 16:1 in France, 11:1 in Germany,10:1 in Japan, and 21:1 in nearby Canada. EvenGreat Britain, with the system of corporate govern-ance most closely similar to that of the USA, hadonly a 25:1 ratio (see Morgenson, 2004). But evenmore important is the shift towards compensatingthe chief executive primarily with stock options.While stock options have come to be widely used inrecent years in Europe, equity compensation consti-tutes a much lower percentage of total CEO com-pensation (even in the UK, it was only 24 per centin 2002) (see Ferrarini et al., 2003, p. 7, fn. 21).European CEOs not only make much less, but theirtotal compensation is also much less performancerelated.14

What explains these differences? Compensationexperts in the USA usually emphasize the tax lawsin the United States, which were amended in theearly 1990s to restrict the corporate deductibility ofhigh cash compensation and thus induced corpora-tions to use equity in preference to cash.15 But thisis only part of the fuller story. Much of the explana-tion is that institutional investors in the USA pres-sured companies for a shift towards equity compen-sation. Why? Institutional investors, who hold themajority of the stock in publicly held companies inthe USA, understand that, in a system of dispersedownership, executive compensation is probably theirmost important tool by which to align managerialincentives with shareholder incentives. Throughoutthe 1960s and 1970s, they had seen senior manage-ments of large corporations manage their firms in arisk-averse and growth-maximizing fashion, retain-ing ‘free cash flow’ to the maximum extent possible.Such a style of management produced the bloated,and inefficient conglomerates of that era (for exam-ple, Gulf & Western and IT&T). Put simply, asystem of exclusively cash compensation createsincentives to avoid risk and bankruptcy and tomaximize the size of the firm, regardless of profit-

13 See Denis et al. (2005). For an earlier study finding that the greater use of option-related compensation results in greater privatesecurities litigation, see Peng and Roell (2004).

14 Ferrarini et al. (2003, pp. 6–7) note that performance-related pay is in wide use only in the UK, and that controlling shareholderstend to resist significant use of incentive compensation.

15 In 1993, Congress enacted Section 162(m) of the Internal Revenue Code, which denies a tax deduction for annual compensationin excess of $1m per year paid to the CEO, or the next four most highly paid officers, unless special tests are satisfied. Its passageforced a shift in the direction of equity compensation. For a fuller account of this change, see Coffee (2004b, pp. 274–5).

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

204

OXFORD REVIEW OF ECONOMIC POLICY, VOL. 21, NO. 2

ability, because a larger firm size generally implieshigher cash compensation for its senior managers.

Once the US tax laws and institutional pressuretogether produced a shift to equity compensation inthe 1990s, managers’ incentives changed, and man-agers sought to maximize share value (as the insti-tutions had wanted). But what the institutions failedto anticipate was that there can be too much of agood thing. Aggressive use of these incentives inturn encouraged the use of manipulative techniquesto maximize stock price over the short run. Althoughsuch spikes may not be sustainable, corporate man-agers possess asymmetric information, and antici-pating their inability to maintain earnings growth,they can exercise their options and bail out.

One measure of this transition is the changing natureof financial irregularities. The Sarbanes–Oxley Actrequired the SEC to study all its enforcement pro-ceedings over the prior 5 years (i.e. 1997–2002) toascertain what kinds of financial and accountingirregularities were the most common. Out of the 227‘enforcement matters’ pursued by the SEC overthis period, the SEC has reported that 126 (or 55 percent) alleged ‘improper revenue recognition’ (SEC,2003, p. 2). Similarly, the earlier noted GAO Studyfound that 38 per cent of all restatements in itssurvey were for revenue recognition timing errors.Either managers were recognizing the next period’srevenues prematurely—or they were simply invent-ing revenues that did not exist. Both forms of errorssuggest that managers were striving to manufacturean artificial (and possibly unsustainable) spike incorporate income.

That managers were systematically able to over-estimate revenues and then recognize them prema-turely in ways that ultimately compelled earningsrestatements shows a market failure—particularlywhen the market penalty for premature revenuerecognition was, as earlier noted, so Draconian as toresult in a 25 per cent decline in market price onaverage (see section II, para. 4). Why did securitiesanalysts accept such optimistic predictions and notdiscount them? Here, the evidence is that very few

analysts downgraded public companies in the monthsprior to earnings restatements—even though shortsellers and insiders had recognized the likelihood ofan earnings restatement.16 Yet, while analysts andauditors may have been slow to recognize prema-ture revenue recognition, considerable evidencesuggests that short sellers were able to recognizethe signals and profit handsomely by anticipatingearnings restatements.17 The implications of thispoint are that smart traders could and did do whatprofessional gatekeepers were insufficiently moti-vated to do: recognize the approach of major marketcollapses. In short, this is a story of ‘gatekeeperfailure’ in that the professional agents of corporategovernance did not adequately serve investors. Inconsequence, the Sarbanes–Oxley Act of 2002understandably focused on gatekeepers and man-dated closer scrutiny of auditors, securities analysts,and credit-rating agencies.

III. FRAUD IN CONCENTRATEDOWNERSHIP REGIMES

The pattern in concentrated ownership systems isvery different, but not necessarily better. In the caseof most European corporations, there is a controllingshareholder or shareholder group.18 Why is thisimportant? A controlling shareholder does not needto rely on indirect mechanisms of control, such asequity compensation or stock options, in order toincentivize management. Rather, it can rely on a‘command and control’ system because, unlike thedispersed shareholders in the USA, it can directlymonitor and replace management. Hence, corpo-rate managers have both less discretion to engage inopportunistic earnings management and less moti-vation to create an earnings spike (because it will notbenefit a management not compensated with stockoptions).

Equally important, the controlling shareholder alsohas much less interest in the day-to-day stock priceof its company. Why? Because the controllingshareholder seldom, if ever, sells its control blockinto the public market. Rather, if it sells at all, it will

16 See Griffin (2002), who reports a study of 847 companies sued in federal securities class actions between 1994 and 2001.17 Desai et al. (2004). For a similar study, see Efendi et al. (2005).18 For excellent overviews of European ownership patterns, see Franks and Mayer (1997), La Porta et al. (1999), and Barca and

Becht (2001).

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

205

J. C. Coffee, Jr

make a privately negotiated sale at a substantialpremium over the market price to an incoming, newcontrolling shareholder. Such control premiums arecharacteristically much higher in Europe than in theUnited States.19 As a result, controlling sharehold-ers in Europe do not obsess over the day-to-daymarket price and rationally do not engage in tacticsto recognize revenues prematurely to spike theirstock price. These two explanations—less use ofequity compensation and less interest in the short-term stock price—explain, at least in part, why therewere fewer accounting irregularities in Europe thanin the USA during the late 1990s.

This generalization may seem subject to counter-examples. For example, some well-known Euro-pean companies—e.g. Vivendi Universal, RoyalAhold, Skandia Insurance, or Adecco20—did expe-rience accounting irregularities. But these are ex-ceptions that prove the rule. Nearly all were USlisted companies whose accounting problems ema-nated from US-based subsidiaries, and several hadtransformed themselves into American-style con-glomerates (the leading example being Vivendi) thateither awarded stock options or needed to maximizetheir short-term stock price in order to make multipleacquisitions.

Potentially, some of this disparity between Europeand the United States could be an artefact of lessrigorous regulatory oversight of public companies inEurope or, alternatively, of the lesser litigation risk inEurope. Hence, European issuers might be lesswilling to restate their financial statements, evenwhen they discover a past error, because they do notexpect regulatory authorities or the plaintiffs’ bar to

hold them accountable. While this point has somevalidity, it should not be overextended. The lessdemanding scrutiny given the financial statementsof public issuers in Europe (particularly in the sec-ondary-market context, where securities are notbeing issued) still does not supply a motor force oran incentive for financial manipulation. Even ifEuropean issuers could inflate their financial state-ments, they had less reason to do so. Finally, finan-cial-statement inflation can lead to the ultimatecollapse of the corporate issuer when the marketeventually discovers the fraud (as Enron andWorldCom illustrate). Here, European examples ofsimilar collapses are conspicuously absent.21

Does this analysis imply that European managersare more ethical, or that European shareholders arebetter off than their American counterparts? By nomeans! Concentrated ownership encourages a dif-ferent type of financial overreaching: the extractionof private benefits of control. Dyck and Zingales(2004) have shown that the private benefits ofcontrol vary significantly across jurisdictions, rang-ing from –4 per cent to +65 per cent, depending insignificant part on the legal protections given minor-ity shareholders.22 While there is evidence that themarket cares about the level of private benefits thatcontrolling shareholders will extract (see Ehrhardtand Nowack, 2003), the market has a relativelyweak capacity to discern on a real-time basis whatbenefits are in fact being expropriated.

In emerging markets, the expropriation of privatebenefits typically occurs through financial transac-tions. Ownership may be diluted through publicofferings, and then a coercive tender offer or

19 See Dyck and Zingales (2004), discussed below; see also Nenova (2000), who finds significantly higher control premiums incountries relying on French civil law and high, but lower, premiums in countries using German civil law; these control premiumswere significantly above the average premiums in common law countries; Zingales (1995).

20 Both the financial scandals at Adecco and Royal Ahold originated in the United States and, at least initially, centred aroundaccounting at US subsidiaries. See Simonian (2004); McCoy (2005) notes that Royal Ahold’s accounting problems began at USFoodservices, Inc., a subsidiary of Royal Ahold, where the US managers were compensated with stock options; Vivendi Universalcan be described as a US-style acquisitions-oriented financial conglomerate. See Johnson (2004).

21 If one goes back far enough, one can certainly find examples of sudden financial collapse in Europe—for example,Metallgesellschaft in 1994. See Fisher (1995). But more recent examples are largely lacking. Also, Metallgesellschaft’s financialdistress seems more to have been the product of the negligent mishandling of derivatives than any fraudulent desire to inflate earnings.See also Edwards and Canter (1995). Such accounting scandals as have occurred in Germany—for example, the fraud at Klockner-Humboldt-Deuta or the collapse of the Jurgen Schneider real estate empire—involved longstanding frauds in which assets wereoverstated and liabilities understated with the apparent acquiescence of both auditors and, sometimes, the principal lending bank.The monitoring failures in these cases much more closely resemble Parmalat than Enron. Many of these failures are reviewed inWenger and Kaserer (1998).

22 See Dyck and Zingales (2004); see also Nenova (2000).

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

206

OXFORD REVIEW OF ECONOMIC POLICY, VOL. 21, NO. 2

squeeze-out merger is used to force minority share-holders to tender at a price below fair market value.These techniques have been discussed in detailelsewhere and in their crudest forms have beengiven the epithet ‘tunnelling’ to describe them.23 Aclassic example was the Bulgarian experience be-tween 1999 and 2002, when roughly two-thirds ofthe 1,040 firms on the Bulgarian stock exchangewere delisted, following freeze-out tender offers forthe minority shares at below market, but still coer-cive, prices.24

In more developed economies, such financial trans-actions may be precluded. Instead, ‘operational’mechanisms can be used: for example, controllingshareholders can compel the company to sell itsoutput to, or buy its raw materials from, a corpora-tion that they independently own. In emerging mar-kets, growing evidence suggests that firms withincorporate groups engage in more related partytransactions than firms that are not members of acontrolled group (see Ming and Wong, 2003). Inessence, these transactions permit controlling share-holders to transfer resources from companies inwhich they have lesser cash-flow rights to ones inwhich they have greater cash-flow rights (seeBertrand et al., 2000).

Although it may be tempting to deem tunnelling andrelated opportunistic practices as characteristic onlyof emerging markets where legal protections arestill evolving, considerable evidence suggests thatsuch practices are also prevalent in more ‘mature’European economies.25 Indeed, some students ofEuropean corporate governance claim that the domi-nant form of concentrated ownership (i.e. absolutemajority ownership) is simply inefficient because itpermits too much predatory misbehaviour (seeKirchmaier and Grant, 2004b).

A danger lurks here in seeking to prove too much.If European corporate governance were as vulner-able to opportunistic behaviour by controlling share-holders as some critics suggest, then one wonderswhy minority shareholders would invest at all andwhy even ‘thin’ securities markets could survive.Perhaps, the answer is that other actors substitutefor the role that gatekeepers play in dispersedownership legal regimes. For example, some arguethat the universal banks, which typically hold large,but non-controlling blocks of stock as well as ad-vancing debt capital, play such a protective role.26

Others point to the impact of co-determination,which gives labour a major voice in corporategovernance in some European countries that it lacksin Anglo-Saxon legal systems.27 Still others point tocross-monitoring by other blockholders.28 All theseanswers encounter problems, which need not beresolved in this essay. All that need be asserted hereis that there is less reason to believe that gatekeep-ers—that is, professional agents serving sharehold-ers but selected by the corporation—work as well inconcentrated ownership regimes as in dispersedownership regimes.

To understand this contention, it is useful to examinethe nature of the scandals that characterize concen-trated ownership systems because they seem toshow a distinct and different type of gatekeeperfailure. Two recent scandals typify this pattern:Parmalat and Hollinger. Parmalat is the paradigm-atic fraud for Europe (just as Enron and WorldComare the representative frauds in the United States).Parmalat’s fraud essentially involved the balancesheet, not the income statement. It failed when a€3.9 billion account with Bank of America proved tobe fictitious.29 At least, $17.4 billion in assets mys-teriously vanished from its balance sheet. Efforts byits trustee to track down these missing funds appear

23 For the article coining this term, see Johnson et al. (2000).24 See Atanasov et al. (2005). These authors estimate that these transactions occurred at about 25 per cent of the shares’ intrinsic value.25 Many commentators have criticized German corporate governance on the grounds that it permits controlling shareholders to

diverge from a pro rata distribution of enterprise cash flows, for example, through one-sided transfer pricing arrangements withaffiliated companies, or asset sales on favourable terms to affiliates. See Gordon (1999). If this is the problem, an independent auditoris probably not the answer, as it cannot stop such transactions. For a more recent overview of the means by which controllingshareholders currently divert value to themselves in Europe, see Kirchmaier and Grant (2004a).

26 This was once the consensus view. But more recently, sceptics have demonstrated that the universal banks in Germany havefew representatives on the supervisory board, tend to do no more monitoring than banks in dispersed ownership regimes, and largelydefer to the managing board’s decisions. See Edwards and Fischer (1994).

27 The impact of co-determination on corporate governance has been much debated. See, for example, Roe (1998).28 See Gorton and Schmid (1996), who suggest the role of non-bank blockholders in monitoring the controlling shareholder.29 This summary of the Parmalat scandal relies upon the Wall Street Journal’s account. See Galloni and Reilly (2004).

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

207

J. C. Coffee, Jr

to have found that at least €2.3 billion were paid toaffiliated persons and shareholders.30 In short, pri-vate benefits appear to have siphoned off to control-ling shareholders through related party transactions.Unlike the short-term stock manipulations that oc-cur in the USA, this was a scandal that had contin-ued for many years, probably for over a decade.

At the heart of the Parmalat fraud lies a crucialfailure by its gatekeepers. Parmalat’s auditors formany years had been an American-based firm,Grant Thornton, whose personnel had auditedParmalat and its subsidiaries since the 1980s (Galloniand Reilly, 2004, p. 10). Although Italian law uniquelymandated the rotation of audit firms, Grant Thorntonfound an easy evasion. It gave up the role of beingauditor to the parent company in the Parmalatfamily, but continued to audit its subsidiaries (Galloniand Reilly, 2004, pp. 9–10). Among these subsidiar-ies was the Cayman-Islands-based subsidiary, BoulatFinancing Corporation, whose books showed thefictitious Bank of America account, the discovery ofwhich triggered Parmalat’s insolvency (Galloni andReilly, 2004, p. 10).

The recent Hollinger scandal also involved over-reaching by controlling shareholders. AlthoughHollinger International is a Delaware corporation,its controlling shareholders were Canadian, as weremost of its shareholders. According to the reportprepared by counsel to its independent directors,former SEC Chairman Richard Breeden, Hollingerwas a ‘kleptocracy’ (see Hollinger, 2004, p. 4). Itscontrolling shareholders allegedly siphoned off morethan $400m from Hollinger—or more than 95 percent of the company’s adjusted net income from1997 to 2003 (Leonard, 2004). On sales of assets byHollinger, its controlling shareholders secretly tooklarge side payments, which they directed be paid tothemselves out of the sales proceeds (Leonard,2004). But bad as the Hollinger case may be, littleevidence suggests that Lord Black and his croniesmanipulated earnings through premature revenuerecognition. What this contrast shows is that con-trolling shareholders may misappropriate assets, buthave much less reason to fabricate earnings. Thisdoes not mean that business ethics are better (or

worse) within a concentrated ownership regime, butonly that the modus operandi for fraud is different.The real conclusion is that different systems ofownership encourage characteristically differentstyles of fraud. The next question becomes whethergatekeepers can play a functionally equivalent pro-tective role in both legal regimes.

IV. GATEKEEPER FAILURE ACROSSOWNERSHIP REGIMES

Both ownership regimes—dispersed and concen-trated—show evidence of gatekeeper failure. TheUSA/UK system of dispersed ownership is vulner-able to gatekeepers not detecting inflated earnings,and concentrated ownership systems fail to theextent that gatekeepers miss (or at least fail toreport) the expropriation of private benefits. A keydifference, of course, is that in dispersed ownershipsystems the villains are managers and the victimsare shareholders, while, in concentrated ownershipsystems, the controlling shareholders overreachminority shareholders.

In turn, this raises the critical issue: can gatekeepersin concentrated ownership systems monitor thecontrolling shareholder who hires (and potentiallycan fire) them? Although there clearly have beennumerous failures by gatekeepers in dispersed own-ership systems, the answer for these systems prob-ably lies in principle in redesigning the governancecircuitry within the public corporation so that thegatekeeper does not report to those that it is expect-ed to monitor. Thus, the auditor or attorney can berequired to report to an independent audit committeerather than corporate managers (as Sarbanes–Oxleymandates). But this same answer does not work aswell in a concentrated ownership system. In such asystem, even an independent audit committee mayserve at the pleasure of a controlling shareholder.

Indeed, some forms of gatekeepers common indispersed ownership systems seem inherently lesslikely to be effective in a system of concentratedownership. For example, the securities analyst isinherently a gatekeeper for dispersed ownership

30 Galloni and Reilly (2004). Parmalat’s former CEO, Mr Tanzi, appears to have acknowledged to Italian prosecutors ‘thatParmalat funneled about Euro 500 Million to companies controlled by the Tanzi family, especially to Parmatour’. See Melis (2004,p. 6). Prosecutors appear to believe that the total diversions to Tanzi family-owned companies were at least €1,500m. Galloniand Reilly (2004, p. 6, n. 2).

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

208

OXFORD REVIEW OF ECONOMIC POLICY, VOL. 21, NO. 2

regimes. In concentrated ownership regimes, thevolume of stock trading in its thinner capital marketsis likely to be insufficient to generate brokeragecommissions sufficient to support a profession ofanalysts covering all publicly held companies. Buteven if analyst coverage in concentrated ownershipregimes were equivalent to that in dispersed owner-ship systems, the analyst’s predictions of the firm’sfuture earnings or value would still mean less topublic shareholders if the controlling shareholderremained in a position to squeeze-out the minorityshareholders.

Even the role of the auditor differs in a concentratedownership system. The existence of a controllingshareholder necessarily affects auditor independ-ence. In a dispersed ownership system, corporatemanagers might sometimes ‘capture’ the audit part-ner of their auditor (as seemingly happened atEnron). But the policy answer was obvious (andSarbanes–Oxley quickly adopted it): rewire theinternal circuitry so that the auditor reported to anindependent audit committee. However, in a con-centrated ownership system, this answer works lesswell because the auditor is still reporting to a boardthat is, itself, potentially subservient to the control-ling shareholder. Thus, the auditor in this system isa monitor who cannot effectively escape the controlof the party that it is expected to monitor. Althoughdiligent auditors could have presumably detectedthe fraud at Parmalat (at least to the extent ofdetecting the fictitious bank account at the CaymanIslands subsidiary), one suspects that they wouldhave likely been dismissed at the point at which theybegan to monitor earnestly. More generally, audi-tors can do little to stop squeeze-out mergers,coercive tender offers, or even unfair related-partytransactions. These require statutory protections ifthe minority’s rights are to be protected. In fairness,shareholders in a concentrated ownership systemmay receive some protection from other gatekeep-ers, including the large banks that typically monitorthe corporation.

There is an important historical dimension to thispoint. The independent auditor arose in Britain in the

middle of the nineteenth century, just as industriali-zation and the growth of railroads was compellingcorporations to market their shares to a broaderaudience of investors (see Littleton, 1988). Amend-ments in 1844 and 1845 to the British CompaniesAct required an annual statutory audit with theauditor being selected by the shareholders.31 Thismade sense, because the auditor was thus placed ina true principal–agent relationship with the share-holders who relied on it. But this same relationshipdoes not exist when the auditor reports to sharehold-ers in a system in which there is a controllingshareholder. Finally, even if the auditor is asked toreport on the fairness of inter-corporate dealings orrelated party transactions, this is not its core compe-tence. Other protections—such as supermajority votes,mandatory bid requirements, or prophylactic rules—may be far more valuable in protecting minorityshareholders when there is a controlling share-holder. This may explain the slower development ofauditing procedures and internal controls in Europe.

Potentially, there is a further implication for the useof gatekeepers in concentrated ownership econo-mies. If the controlling shareholder can potentiallydominate the selection of the auditor or other gate-keepers, then it becomes at least arguable that if theauditor is to serve as an effective reputationalintermediary, it should be selected by the minorityshareholders and report to them. This article doesnot attempt to design such an unprecedented sys-tem, but smugly contents itself with pointing out thelikely inadequacy of alternative systems. The sec-ond-best alternative would appear to be accordingthe auditor’s selection, retention, and compensationto the independent directors.

V. CONCLUSION

This article’s generalizations are not presented asiron laws. ‘Private benefits of control’ can bemisappropriated in a US public company, and recentillustrations include the Adelphia scandal, where acontrolling family diverted assets of over $3 billion toitself in much the same way as did the controlling

31 The 1844 amendment was to the Joint Stock Companies Act (see 7 & 8 Vict., Ch. 110) (1844), and the 1845 amendment wasto the Companies Clauses Consolidation Act (see 8 & 9 Vict., Ch. 16) (1845). For a more detailed review of this legislation, seeO’Conner (2004).

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

209

J. C. Coffee, Jr

REFERENCES

Anderson, K. L., and Yohn, T. L. (2002), ‘The Effect of 10K Restatements on Firm Value, Information Asymmetries,and Investors’ Reliance on Earnings’, available at http://ssrn.com/abstract=332380

Atanasov, V., Ciccotello, C. S., and Gyoshev, S. B. (2005), ‘How Does Law Affect Finance? An Empirical Examinationof Tunneling in an Emerging Market’, William Davidson Institute Working Paper No. 742, available at http://ssrn.com/ abstract=423506

Banner, S. (1997), ‘What Causes New Securities Regulation? 300 Years of Evidence’, Washington University LawQuarterly, 75(2), 849.

Barca, F., and Becht, M. (eds) (2001), The Control of Corporate Europe, Oxford, Oxford University Press.Bertrand, M., Mehta, P., and Mullainathan, S. (2000), ‘Ferreting Out Tunneling: An Application to Indian Business

Groups’, MIT Dept of Economics Working Paper No. 00-28, available at http://ssrn.com/abstract=246001Cheng, Q., and Warfield, T. (2004), ‘Equity Incentives and Earnings Management’, available at http://ssrn.com/

abstract=457840Coffee, J. C., Jr. (2001), ‘The Rise of Dispersed Ownership: The Roles of Law and the State in the Separation of Ownership

and Control’, Yale Law Journal, 111(1), 1–82. — (2004a), ‘Gatekeeper Failure and Reform: The Challenge of Fashioning Relevant Reforms’, Boston University

Law Review, 84(2), 301–64. — (2004b), ‘What Caused Enron?: A Capsule Social and Economic History of the 1990s’, Cornell Law Review,

89(2).Denis, D. J., Hanouna, P., and Sarin, A. (2005), ‘Is There a Dark Side to Incentive Compensation?’, available at http:/

/ssrn.com/abstract=695583Desai, H., Krishnamurthy, S., and Venkataraman, K. (2004), ‘Do Short Sellers Target Firms with Poor Earnings Quality?:

Evidence from Earnings Restatements’, available at http://ssrn.com/abstract=633283Dyck, A., and Zingales, L. (2004), ‘Private Benefits of Control: An International Comparison’, Journal of Finance, 59(2),

537–600.Edwards, F., and Canter, M. (1995), ‘The Collapse of Metallgesellschaft: Unhedgeable Risks, Poor Hedging Strategy,

or Just Bad Luck’, Journal of Futures Markets, 15(3), 211–64.Edwards, J., and Fischer, K. (1994), Banks, Finance and Investment in Germany, Cambridge, Cambridge University

Press.Efendi, J., Kinney, M. R., and Swanson, E. P. (2005), ‘Can Short Sellers Predict Accounting Restatements?’, AAA 2005

FARS Meeting Paper, available at http://ssrn.com/abstract= 591361 — Srivastava, A., and Swanson, E. P., (2004), ‘Why Do Corporate Managers Misstate Financial Statements: The

Role of Option Compensation, Corporate Governance and Other Factors’, available at http://ssrn.com/abstract=547920

Ehrhardt, O., and Nowak, E. (2003), ‘Private Benefits and Minority Shareholder Expropriation (or What Exactly ArePrivate Benefits of Control)’, EFA Annual Conference Paper No. 809, available at http://ssrn.com/ab-stract=423506

shareholders at Parmalat.32 Similarly, Europeancompanies with dispersed ownership have en-gaged in US-style earnings management.33 Thepoint then is that, across different legal regimes,the structure of share ownership likely deter-mines the nature of the fraud. The bottom line forregulators is this: show us the structure of shareownership, and we can predict the likely scenariofor fraud and abuse.

Public policy thus needs to start from the recognitionthat dispersed ownership creates managerial incen-tives to manipulate income, while concentratedownership invites the low-visibility extraction ofprivate benefits. As a result, governance protec-tions that work in one system may fail in the other.Even more importantly, different gatekeepers needto be designed into different governance systems tomonitor for different abuses.

32 Adelphia Communications went into bankruptcy after it was discovered that it had made undisclosed loans and loan guaranteesof at least $3.1 billion to its founder’s family. This closely resembles Parmalat in that the fraud involved both the balance sheetand diversions to the founder’s family. In bankruptcy, Adelphia sued its former auditors, Deloitte & Touche, for allegedly failingto disclose or investigate Adelphia’s loans to the Rigas family. See Frank (2002).

33 For examples, see section III, para. 3.

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

210

OXFORD REVIEW OF ECONOMIC POLICY, VOL. 21, NO. 2

Ferrarini, G., Moloney, N., and Vespro, C. (2003), ‘Executive Remuneration in the EU: Comparative Law and Practice’,ECGI Working Paper.

Fisher, A. (1995), ‘Metallgesellschaft: The Oil Deals That Crippled a German Metal-trading Giant’, Financial Times,20 March, 20.

FitchRatings (2005), ‘Accounting and Financial Reporting Risk: 2005 Global Outlook’, 14 March.Frank, R. (2002), ‘Adelphia Sues Deloitte & Touche, Accusing Former Auditor of Fraud’, Wall Street Journal, 7

November, A2.Franks, J., and Mayer, C. P. (1997), ‘Corporate Ownership and Control in German, the UK and France’, Journal of

Applied Corporate Finance, 9(4), 30.Galloni, A., and Reilly, D. (2004), ‘How Parmalat Spent and Spent’, Wall Street Journal, 23 July.Gordon, J. N. (1999), ‘Pathways to Corporate Convergence?: Two Steps On the Road to Shareholder Capitalism in

Germany’, Columbia Journal of European Law, 5(219).Gorton, G., and Schmid, F. A. (1996), ‘Universal Banking and the Performance of German Firms’, National Bureau of

Economic Research Working Paper No. 5453.Griffin, P. A. (2002), ‘A League of Their Own? Financial Analysts’ Responses to Restatements and Corrective

Disclosures’, available at http://ssrn.com/abstract= 326581Hall, B. J. (2003), ‘Six Challenges in Designing Equity-based Pay’, Accenture Journal of Applied Corporate Finance,

15(3), 21.Hollinger International (2004), ‘Report of Investigation by the Special Committee of the Board of Directors of Hollinger

International Inc’, 30 August.Holmstrom, B., and Kaplan, S. (2003), ‘The State of US Corporate Governance: What’s Right and What’s Wrong’,

Accenture Journal of Applied Corporate Finance, 15(3), 8.Huron Consulting Group (2003a), An Analysis of Restatement Matters: Rules, Errors, Ethics, for the Five Years Ended

December 31, 2002. — (2003b), 2003 Annual Review of Financial Reporting Matters. — (2005), ‘New Report by Huron Consulting Group Reveals Financial Restatements Increased at a Record Level

in 2004’, Press Release, 19 January.Johnson, J. (2004), ‘Vivendi Probe Ends With Euro $1 Million Fine for Messier’, Financial Times, 8 December, 31.Johnson, S. A., Ryan, H. E., Jr, and Tian, Y. S. (2003), ‘Executive Compensation and Corporate Fraud’, available

at http://ssrn.com/abstract=395960 — La Porta, R., Lopez-de-Silanes, F., and Shleifer, A. (2000), ‘Tunneling’, American Economic Review, 90(2), 22–7.Kirchmaier, T., and Grant, J. (2004a), ‘Financial Tunneling and the Revenge of the Insider System: How to Circumvent

the New European Corporate Governance Legislation’, available at http://ssrn.com/abstract=613945 — — (2004b), ‘Corporate Ownership Structure and Performance in Europe’, CEP Discussion Paper No. 0631,

available at http://ssrn.com/ abstract=616201La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R. W. (1999), ‘Corporate Ownership Around the World’,

Journal of Finance, 54(2), 471–517.Leonard, D. (2004), ‘More Trials for Lord Black’, Fortune, 4 October, 42.Littleton, A. C. (1988), Accounting Evolution to 1900, University of Alabama Press.McCoy, K. (2005), ‘Prosecutors Charge Nine More in Royal Ahold Case’, USA Today, 14 January, 1B.Melis, A. (2004), ‘Corporate Governance Failures. To What Extent is Parmalat a Particularly Italian Case?’, available

at http://ssrn.com/abstract=563223Ming, J. J., and Wong, T. J. (2003), ‘Earnings Management and Tunneling through Related Party Transactions:

Evidence from Chinese Corporate Groups’, EAA 2003 Annual Conference Paper No. 549, available at http://ssrn.com/abstract=42488

Morgenson,G. (2004), ‘Explaining (Or Not) Why the Boss Is Paid So Much’, New York Times, 25 January, §3, 1.Moriarty, G. B., and Livingston, P. B. (2001), ‘Quantitative Measures of the Quality of Financial Reporting’, Financial

Executive, July/August, 53–4.Nenova, T. (2000), ‘The Value of Corporate Votes and Control Benefits: A Cross-country Analysis’, available

at http://ssrn.com/abstract=237809O’Conner, S. M. (2004), ‘Be Careful What You Wish For: How Accountants and Congress Created the Problem of

Auditor Independence’, Boston College Law Review, 45(4), 741–828.Peng, L., and Roell, A. (2004), ‘Executive Pay, Earnings Manipulation and Shareholder Litigation’, AFA Philadelphia

Meetings, December, available at http://ssrn.com/ abstract=488148Richardson, S., Tuna, I., and Wu, M. (2002), ‘Predicting Earnings Management: The Case of Earnings Restatements’,

available at http://ssrn.com/abstract=338681

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from

211

J. C. Coffee, Jr

Roe, M. J. (1998), ‘German Co-determination and German Securities Markets’, Columbia Business Law Review, 1, 167.SEC (2003), Report Pursuant to Section 704 of the Sarbanes–Oxley Act of 2002, Washington, DC, Securities and

Exchange Commission.Simonian, H. (2004), ‘Europe’s First Victim of Sarbanes–Oxley? Corporate Governance: Adecco was Pilloried for its

Poor Handling of US Accounting Problems’, Financial Times, 29 January, 14.GAO (2002), Financial Statement Restatements: Trends, Market Impacts, Regulatory Responses and Remaining

Challenges, Washington, DC, US General Accounting Office, Pub. No. 03-138.Wenger, E., and Kaserer, C. (1998), ‘German Banks and Corporate Governance—A Critical View’, in K. Hopt et al. (eds),

Comparative Corporate Governance—The State of the Art and Emerging Research, Oxford, Clarendon Press.Zingales, L. (1995), ‘What Determines the Value of Corporate Votes?’, Quarterly Journal of Economics, 110(4), 1047–

73.

at Hong K

ong Polytechnic U

niversity on May 18, 2010

http://oxrep.oxfordjournals.orgD

ownloaded from