10 accounting theory questions.

Accounting Horizons American Accounting Association Vol. 28, No. 3 DOI: 10.2308/acch-50759 2014 pp. 627–671

COMMENTARY

SOX after Ten Years: A Multidisciplinary Review

John C. Coates and Suraj Srinivasan

SYNOPSIS: We review and assess research findings from more than 120 papers in accounting, finance, and law to evaluate the impact of the Sarbanes-Oxley Act. We

describe significant developments in how the Act was implemented and find that despite

severe criticism, the Act and institutions it created have survived almost intact since

enactment. We report survey findings from informed parties that suggest that the Act has

produced financial reporting benefits. While the direct costs of the Act were substantial

and fell disproportionately on smaller companies, costs have fallen over time and in

response to changes in its implementation. Research about indirect costs such as loss of

risk taking in the U.S. is inconclusive. The evidence for and social welfare implications of

claimed effects such as fewer IPOs or loss of foreign listings are unclear. Financial

reporting quality appears to have gone up after SOX but research on causal attribution is

weak. On balance, research on the Act’s net social welfare remains inconclusive. We

end by outlining challenges facing research in this area, and propose an agenda for

better modeling costs and benefits of financial regulation.

Keywords: Sarbanes-Oxley; regulation; SEC; PCAOB; audit; cost benefit analysis.

INTRODUCTION

I n this article we review and assess over 120 studies of the Sarbanes-Oxley Act, focusing on

research in accounting, law, and finance after 2005. We describe major developments in its

legal, regulatory, and institutional implementation; its effects on U.S. corporate law, disclosure

practices, and other countries’ laws; and the propensity of companies to go or remain public in the

U.S. We also note a puzzle regarding the Act’s reception in public debate. On the one hand, the law

continues to be fiercely and relentlessly attacked in the U.S., particularly in political election battles

John C. Coates is a Professor and Suraj Srinivasan is an Associate Professor, both at Harvard University.

We thank participants at the Harvard Business School Information, Markets, and Organizations Conference; workshop participants from the HBS Finance unit; the reviewers, Jeffrey Burks, James Cox, Peter Iliev, Michael Klausner, Clive Lennox, Christian Leuz, Lynn Paine, Krishna Palepu, Shivaram Rajgopal, Mark Roe, Eugene Soltes, Urska Velikonja, and Paul Zarowin for their valuable comments, and Paul A. Griffin (editor) for encouraging us to write this paper.

Submitted: October 2013 Accepted: January 2014

Published Online: September 2014 Corresponding Author: Suraj Srinivasan

Email: [email protected]

627

and during legislative debates, reflected in part in provisions of the Dodd-Frank Act and the JOBS

Act, which can be seen as a partial legislative rollback of the Act. On the other hand, we discuss

survey evidence that suggests that informed observers, including corporate officers and investors,

do not believe that the Act—as implemented, taking into account significant relaxations of its most criticized provision (section 404[b] internal control attestation)—has been a significant problem,

and may well have produced net benefits, and the law has been copied at least in part by other

countries. What explains this puzzle of continued hostility amid acquiescence or even mild praise

by those most directly affected by the Act?

We would have liked to have then evaluated the Act itself, at least provisionally, but as we

explain in our review of the literature, the state of research is such that—even after ten years—no

conclusions can be drawn about the net costs and benefits of the Act, its effects on net shareholder

wealth, or other research relevant to its assessment. We suggest that the puzzle of the Act’s

reception may be explained in part by the inconclusive state of research on its costs and benefits,

and so follows a two-part pattern that may be generalized to most types of regulation in the spheres

of financial institutions, markets, and corporate governance. First—as with most types of

regulation—the Act had clear, non-trivial, and quantifiable direct costs. Second, the tasks of

estimating either the benefits or the indirect costs of the Act are at least an order of magnitude more

difficult than the task of estimating direct costs, and are possibly beyond the present capacity of

researchers to achieve with much precision.

Absent research designs adequate to the task of drawing even provisionally reliable causal

inferences about the Act’s full range of effects, the prior beliefs of researchers about its net

benefits—which range from large and positive to large and negative—remain largely unaffected by

the research that has been done. Absent scholarly consensus, political entrepreneurs have used clear

(if overstated) evidence on direct costs to deride the Act as a symbol of regulatory overreach,

despite the view among informed observers that the costs and benefits of the law as implemented have been at worst roughly equivalent and possibly net positive.

The stakes implicit in the ongoing uncertainty about the Sarbanes-Oxley Act are large, with

lessons for policy-makers and researchers. We sketch a research agenda in light of our literature

review and assessment. For researchers, the challenges are to develop methods that better specify

and estimate (or at least bound) the Act’s potential benefits, as well as its indirect costs, and we

preliminarily suggest ways forward on these tasks. For policy-makers, the challenges are how to

better design future regulatory (or de- or re-regulatory) interventions so as to permit more reliable

inferences about their effects, to improve the quality of information about whether they have led to

net benefits, and to reduce the risk that pure politics, untethered by fact or reason, will continue to

generate unnecessarily costly oscillation in systemically important laws, regulations, and

institutions that form the foundations of capitalism. These tasks are all the more important given

ongoing efforts to legally mandate quantified cost-benefit analysis of financial regulation, which our

findings here suggest remains aspirational, rather than feasible. 1

SARBANES-OXLEY ACT REMAINS LARGELY INTACT, WITH SIGNIFICANT MODIFICATIONS TO SECTION 404’S IMPLEMENTATION

The Sarbanes-Oxley Act (SOX or the Act) was intended to improve auditing of U.S. public

companies, consistent with the law’s official name—The Public Company Accounting Reform and Investor Protection Act of 2002. The Act had two core goals: (1) to create a quasi-public institution to oversee and regulate auditing (the Public Company Accounting Oversight Board [PCAOB]), and

(2) to enlist auditors more extensively in the enforcement of existing laws against theft and fraud by

1 For discussion of cost-benefit analysis and legal efforts to mandate it, see Coates (2014).

628 Coates and Srinivasan

Accounting Horizons September 2014

corporate officers, pursuant to regulations from and enforcement by PCAOB. Reinforcing this core

were new rules concerning the relationships between public companies and their auditors.

PCAOB Is a Durable and Significant Part of the Regulatory Landscape

Prior to SOX, auditors were subject only to state licensing and lightly enforced self-regulation.

From 1978, a part-time ‘‘Public Oversight Board’’ oversaw auditors, but was dominated by audit firms and had little authority. Recognizing that Congress was in no position to legislate auditing

standards in detail, Congress delegated that task in the Act, not to an existing or new public

regulatory body, such as the Securities and Exchange Commission (SEC), nor to leave it in the

private hands of the American Institute of Certified Public Accountants (AICPA), but instead to

create a unique new hybrid public/private agency in the form of the PCAOB.

PCAOB is a non-profit private corporation charged with the public function of overseeing

auditors of SEC-registered companies with the goals of protecting investors and the public interest

in the ‘‘preparation of informative, fair, and independent audit reports.’’ PCAOB’s main tasks are to register, set standards for, inspect, investigate, and discipline public company audit firms. Two, but

only two, of its five full-time directors are auditors, and all of its directors serve staggered five-year

terms after being appointed by the SEC. The SEC must approve the PCAOB’s budget and

regulations, but PCAOB is empowered to directly tax (formally, impose fees on) public companies

and audit firms. As a result, its funding is sheltered from the normal Congressional budget process.

As a private organization, PCAOB is exempt from so-called ‘‘sunshine’’ laws and can operate largely without the constant pressure of public scrutiny that applies to the SEC and other public

bodies. PCAOB can set and enforce regulations; it enjoys broad immunity from private lawsuits;

and its communications are sheltered behind a regulatory privilege, making them generally not

subject to ordinary discovery in lawsuits against audit firms.

This hybrid public/private structure led to a high-profile lawsuit challenging the PCAOB as

unconstitutionally invading the sphere of the executive branch (Free Enterprise Fund et al. v. Public Company Accounting Oversight Board et al.). In 2010, the U.S. Supreme Court largely upheld the agency’s design and structure, striking down only a provision of the Act that limited

removal of PCAOB directors to situations of misbehavior (‘‘for cause’’), increasing the theoretical power of the SEC over the PCAOB. A broader effort to overturn SOX altogether in the same

lawsuit was turned back, however, and PCAOB is now a durable part of the regulatory landscape,

largely intact despite frequent political attacks on SOX itself. 2

PCAOB has a staff of over 600 and

an annual budget of $180 million. Its budget has grown significantly since 2002, both absolutely

and relative to benchmarks such as the SEC’s budget—PCAOB’s budget is now roughly one sixth

of the SEC’s budget, as illustrated in Figure 1, Panel A. By contrast, while the SEC received a large

boost in its own budget in 2003, partly as a result of SOX, its budget has grown more slowly,

roughly one-third the PCAOB growth rate since 2003 (see Figure 1, Panel B).

The Audit Firm/Issuer Relationship Remains as SOX Left It

Just as the structure of PCAOB remains largely unchanged since SOX, so too does the shape of

another major focus of SOX: the relationships between audit firms and public company issuers. In

2 One minor change in the design of PCAOB was adopted in the Dodd-Frank Act in 2010: the PCAOB was required to create and populate an investor advisory board, similar to (and in fact overlapping in membership with) the SEC’s advisory committee, also created by that law. On the advisory board are representatives of large institutional advisors such as CalPERS, TIAA-CREF, Vanguard, and the AFL-CIO; accounting academics; corporate governance professionals; and retired judges and regulatory officials. While the advisory board has no formal power, it represents a direct channel of communication from the investor community to PCAOB, and may further buffer the influence of the audit industry over PCAOB.

SOX after Ten Years: A Multidisciplinary Review 629

Accounting Horizons September 2014

SOX, as well in contemporaneous changes in stock exchange listing standards, those relationships

were significantly changed. Major changes were made (1) to increase the independence of audit

committees of public companies, (2) to transfer control of the audit firm relationship to audit

committees (from management and the board as a whole), and (3) to give audit committees plenary

authority to hire and pay audit firms and other advisors needed for such committees to conduct their

now-statutorily mandated responsibilities. These changes represented a significant change to basic

corporate law and governance, which traditionally had been determined at the state (not federal)

level, and which traditionally had given the full board of directors authority to delegate whatever

authority over the audit firm relationship as it saw fit (including to the management of a public

company). In addition, SOX mandated audit partner rotation every five years, and banned many

non-audit services that had formed the backbone of the consulting businesses that each of the large

audit firms had developed alongside their traditional audit lines of business.

While SOX prohibited audit firms from providing many (but not all) types of consulting

services to their audit clients, just slightly preceding SOX, four of the then Big 5 audit firms spun

off their consulting arms into separate entities in 2000 and 2001, Deloitte being the only one to not

do so. Despite this drastic change in business models arising from the spin-offs, consulting services

still account for a large share of revenues for the big audit firms. Non-audit fees to audit clients as a

proportion of total fees fell from almost 51 percent of fees in 2002 to about 21 percent in 2005, and

have remained steady at that level since then until recently, and were around 22 percent in 2012

FIGURE 1 PCAOB and SEC Budgets

Panel A: Level of PCAOB Budget over Time (Bars) and the Growth in PCAOB Budget Relative to the SEC Budget

(continued on next page)

630 Coates and Srinivasan

Accounting Horizons September 2014

(Whalen, McCoy, and Cross 2013). But these services are now largely provided to companies that

are not audit clients.

Few changes have been made to the rules governing the relationships between audit firms and

public company issuers since SOX. In 2003, the U.S. Government Accountability Office (GAO)

released a study on audit firm rotation (as opposed to audit partner rotation), but little was done to pursue further reform until 2012, when the PCAOB issued a concept release on the topic, prompted

in part by proposals by the European Community in December 2011 to require audit firm rotation

every six years. The PCAOB release generated 630 public comments, and attracted some support in

the academic community (e.g., Bazerman and Moore 2011). However, as of this writing, neither the

PCAOB nor the SEC has followed up on the idea. On July 8, 2013, the House of Representatives

passed H.R. 1564, the Audit Integrity and Job Protection Act, which would amend SOX to prohibit PCAOB from requiring auditor rotation, but as of this writing, the Senate has not taken up the bill.

Internal Control Attestation (Section 404(b)) Has Been Significantly Modified in the Course of Implementation

The other core component of SOX—the mandate that public companies obtain audit firm

attestation over their internal control (IC) systems—proved to be the most controversial. In general

terms, SOX required companies to pay audit firms for what were initially costly, time-consuming,

detailed, and what many viewed as unjustified reviews of companies’ policies, procedures, and

FIGURE 1 (continued)

Panel B: Level of the SEC Budget over Time

The y-axis in Panel B is in million dollars.

SOX after Ten Years: A Multidisciplinary Review 631

Accounting Horizons September 2014

technologies for preventing theft and fraud. These requirements produced sufficiently loud,

widespread, and sustained complaints that PCAOB made significant changes in how SOX section

404 was implemented. Even there, however, the requirement of attestation remains intact for most

public companies, and neither Congress nor PCAOB has adopted major changes to the attestation

process since 2007.

Contrary to popular impression, SOX did not mandate any particular change in companies’

control systems. In 1977, the Foreign Corrupt Practices Act required public companies to have an

IC system that would provide ‘‘reasonable assurances’’ that ‘‘transactions are recorded as necessary to permit preparation of financial statements in conformity’’ with GAAP, and nothing in SOX changed this requirement. Even after SOX, a public company can, if it chooses, adopt any IC

system it believes meets that requirement—even one with what its auditors believe are significant

weaknesses.

What SOX section 404 did was to force disclosure of IC weaknesses, and then to rely on

pressures that flow from such disclosures to cause companies to improve their systems. 3

Only if

market forces or litigation risks are sufficiently powerful do companies actually need to change their

behavior, in the form of investments to or fees related to their IC systems. An open question at the

time SOX was adopted was the extent to which firms would incur all expenses necessary to avoid

negative statements about IC weaknesses, or whether they allow weaknesses to exist and persist,

despite possible negative market reactions to the disclosures of those weaknesses required by SOX.

As discussed more in the section on SOX’s costs, SOX-mandated disclosures did induce

significant direct costs. Partly because of those costs, and the criticism that followed, the SEC,

PCAOB, and Congress have taken further post-SOX actions in response. The SEC deferred

implementation of section 404 for companies with market capitalizations of less than $75 million,

and extended that deferral several times, until 2010, when Congress made it permanent in the Dodd-

Frank Act. In 2006, the SEC adopted a rule permitting firms to defer implementation of section 404

for up to two years after going public, which Congress extended to up to five years in 2012 for all

but the largest newly public companies (i.e., those with market capitalizations under $700 million

and revenues and non-convertible debt under $1 billion).

For firms subject to section 404, the PCAOB in 2007 adopted Audit Standard 5 (AS5), which

significantly relaxed the attestation requirements from those initially adopted in Audit Standard 2

(AS2) in 2004. Among other things, the 2007 changes permit a unified audit and attestation process;

top-down risk-based approaches that permit audit firms to focus on key control risks; the use of

‘‘scaled’’ approaches to take account of firm or IC system component size; the ability to rely on the work of others in the attestation process, so less expensive employees or third party vendors can

directly perform some of the work required; and the replacement of time-consuming and elaborate

‘‘walk-throughs’’ with other methods of testing IC systems. In a large SEC survey conducted after these changes, most respondents reported that these

changes had been economically meaningful, reducing costs by 25 percent or more per year (SEC

2009). Doogar, Sivadasan, and Solomon (2010) find that, relative to the initial AS2 benchmark,

AS5 audit fees became lower on average for all clients and better aligned with client fraud risk (i.e.,

lower for lower fraud risk clients and higher for clients with higher fraud risk). As implemented

then, SOX’s requirements of IC disclosures and attestation have been significantly loosened, both

as to what companies are covered and as to what the Act requires. Any fair assessment of SOX as

3 SOX section 302 likewise imposes only disclosure obligations on CEOs and CFOs, requiring them to certify that a company’s SEC filings do not contain material misstatements or omissions; that the officers have evaluated their companies’ control systems, which have been designed to ensure material information is disclosed to the officers; and that they have disclosed to the audit committee all significant deficiencies or material weaknesses in the control system.

632 Coates and Srinivasan

Accounting Horizons September 2014

initially enacted would acknowledge that the possibility of such loosening in response to pushback

during implementation was part of Congress’s original design, because it gave the PCAOB power

to relax its initial requirements precisely in this way. We return to this point below, as it may partly

explain divergent assessments of SOX.

QUALITATIVE EVIDENCE OF SOX’S IMPACTS

SOX and the rules adopted under it by the SEC and PCAOB have now had ten years to shape

behavior. Before we review studies that try to measure SOX’s costs and benefits, this section

reviews qualitative evidence on some of the specific effects that were predicted as consequences of

SOX, including: (1) whether SOX ‘‘federalized’’ corporate law, as commonly asserted by its critics; (2) whether SOX has functioned as a ‘‘command and control’’ law, as many have assumed, or more in the nature of a ‘‘comply or explain’’ disclosure-forcing law, often asserted to be more true of similar laws in other countries; and (3) whether other countries have followed or repudiated the

U.S. innovations in SOX. Our general conclusions are that SOX had little impact on the federal/

state balance of legal authority over corporations; has functioned to force disclosure, which in turn

has combined with market forces to induce significant changes in control systems; and has been

partly but not completely copied by other countries.

Lack of Federalization of Corporate Law

One prominent criticism immediately following SOX was that it ‘‘federalized’’ corporate law (e.g., Romano 2005; Butler and Ribstein 2006 ). Straightforwardly, the law did not do this. Other

than through the changes in the audit committee/audit firm relationship discussed above, and a

relatively toothless ban on loans to public company executive officers, SOX did little directly to

alter ‘‘corporate governance’’ or the state laws and stock exchange listing standards that shape governance practices. While major changes in corporate governance did coincide with SOX, they

were the result of changes in listing standards adopted by stock exchanges themselves (see Coates

[2007, 109–112] for more discussion).

However, it remained possible that SOX would have powerful effects on corporate governance

in two indirect ways. First, it might pave the way for more Congressional intervention, representing

a step on a slippery slope to a series of federal legislative changes to corporate law and governance

(e.g., Butler and Ribstein 2006 ). Second, it might result in changes in state law as a result of

shareholders suing under state law but using SOX’s requirements as a basis for doing so, with the

result that SOX would effectively create new standards of conduct for corporate directors (e.g.,

Burch 2006; Ferola 2007; Jones 2004).

Ten years later, it is apparent that neither of these possibilities has come about. While the

Dodd-Frank Act may represent a modest example of the first possible indirect effect, in that it

contained general corporate governance provisions that went beyond the financial industry, those

changes are even more modest than those in SOX itself. ‘‘Say on pay’’ is perhaps the most well- known part of Dodd-Frank’s corporate governance provisions, but that requirement merely imposes

a non-binding shareholder vote on executive compensation, and fewer than two percent of U.S.

public companies have experienced a negative vote since its implementation. Other aspects of

Dodd-Frank fall squarely into the ‘‘comply or explain’’ category, such as a requirement that public companies disclose whether they have two different individuals serving as chairman of the board

and CEO, or are extremely modest in their effects, with its authorization (but not a requirement) for

the SEC to adopt regulations permitting shareholders to nominate one to three directors in the

company’s proxy statement (allowing economization on costs for nomination of a minority of

directors), a step that Kahan and Rock (2011) call ‘‘insignificant.’’ Further, the most significant set of changes to U.S. securities laws since SOX can be found not in Dodd-Frank, but in the JOBS Act,

SOX after Ten Years: A Multidisciplinary Review 633

Accounting Horizons September 2014

a 2012 law that mandated the SEC adopt a number of new exemptions from the disclosure and

registration requirements of the federal securities laws—precisely in the opposite direction that

SOX critics predicted. Indeed, by expanding exemptions for widely traded but unlisted firms, the

JOBS Act greatly cut back the scope of SOX’s coverage.

Turning to state law, a comprehensive review of all Delaware court decisions (n¼1,293) in the period from 2002 to 2012 shows that only 15 referred in any way to SOX or its provisions. Of

those, not one imposed liability on directors for failing to adhere to standards or live up to

obligations created by SOX. In fact, no such decision even allowed such a claim to go to trial. Nor

has there been a material change in the formal standards of Delaware case law (for example by

tightening the definition of what it means to be an ‘‘independent’’ director, as asserted by Ferola

[2007]). 4

Nor is there any evident pattern in outcomes of decisions holding corporate directors

personally liable (see Black, Cheffins, and Klausner 2006 ), or in a significantly higher risk of

liability for officers, despite section 302 of the Act (see Vogel 2009). Broad claims that SOX has

distorted Delaware case law turn out to be overstated at best. Roe (2009) may be right that the fact

or threat of lawmaking by Congress or the SEC influences Delaware judges, but the influences are

subtle and hard to see in the case law overall. It may be more accurately understood as both federal

and state lawmakers reacting simultaneously to time-varying public pressures and political demand

for corporate accountability.

In sum, whatever the effects of SOX more broadly, it has not had the effect of ‘‘federalizing’’

corporate law to any meaningful extent. Delaware (and other state) laws, together with stock

exchange listing standards, remain the core of corporate governance in the U.S., even for public

companies. (We discuss data on SOX’s effect on federal securities litigation, as opposed to state

corporate litigation, below.)

SOX’s ‘‘Comply or Explain’’ Features in Operation

Another criticism of SOX was that it was excessively ‘‘mandatory,’’ directly requiring changes in

business decisions, contrary to the tradition of U.S. federal securities laws’ focus on disclosure, which

permitted business decisions to be made in light of market forces. As discussed above, in one key

respect this criticism of SOX was misguided—section 404 of SOX, which requires attestation of IC

systems, is effectively a ‘‘comply or explain’’ regime, in that it permits companies to allow IC systems

to contain weaknesses, as long as that fact is disclosed as part of their own disclosures (and by the

audit firm providing the attestation). Nevertheless, it was possible that market reactions and litigation

risks associated with IC system weaknesses would be such that no public company would allow them

to persist, once identified by an audit firm (even if company officials disagreed with the audit firm).

Indeed, Hammersley, Myers, and Shakespeare (2008) document a market reaction of negative 0.95

percent for material weakness disclosures, suggesting that the material weakness disclosure reports

have value-relevant information. Ashbaugh-Skaife, Collins, Kinney, and LaFond (2009) report a 93-

basis-point increase in cost of equity around the first disclosure of an IC deficiency. Further, firms that

remediate IC deficiencies benefit by a 151-point decrease in cost of equity. Most interestingly,

Ashbaugh-Skaife et al. (2009) report that firms with characteristics that made them most likely to

report IC problems, but which did not report such problems (instead reporting an unqualified SOX

404 audit opinion), experienced a 116-basis-point decrease in cost of equity.

4 Compare In re: The Limited, 2002 Del. Ch. LEXIS 28 (Del. Ch. Mar. 27, 2002) and In re JPMorgan Chase & Co. Shareholder Litigation, 906 A.2d 2005 (Del. Ch. 2005) ( pre- and post-SOX cases, both citing Aronson v. Lewis, 473 A.2d 805, 814 (Del. 1984), both defining ‘‘disinterested and independent’’ director in nearly identical terms and both finding a director ‘‘independent’’ despite being an executive of a company with significant business relations with the director’s company).

634 Coates and Srinivasan

Accounting Horizons September 2014

However, Johnstone, Li, and Rupley (2011) show that many U.S. public companies disclose

persistent IC weaknesses. In the three years ending 2006, they found that 733 U.S. companies

disclosed a material weakness. By the end of year one, most (59 percent) had remedied the

weakness, consistent with market forces and litigation risk pressuring companies to eliminate such

weaknesses by incurring IC costs. Even after three years, nearly a third (30 percent) continued to

disclose the same IC weakness.

Related evidence on this comes from Rice and Weber (2012). They find a significant

proportion of firms fail to report material weaknesses when they exist. Only 32.4 percent of firms

that subsequently made a material restatement previously reported a material weakness. That is,

most firms did not report the IC weakness when it existed, instead reporting the weakness after a more serious restatement had occurred. (Not all restatements result from control weaknesses, but

material restatements often trigger re-evaluation of a firm’s control systems by both the firm and its

auditors.) This suggests that for a significant number of public companies, SOX’s section 404 has

functioned at least in part in a ‘‘comply or explain’’ fashion, contrary to strong characterizations of

that part of the law as ‘‘mandating’’ corporate governance changes.

Further Rice, Weber, and Wu (2013) show weak incentives for timely reporting of section 404

weaknesses. They find firms that do not report a timely IC weakness and later have a restatement

are in fact less likely to have class action lawsuits, SEC sanctions, and management and auditor

turnover compared to firms with restatement where the IC weakness had been previously reported.

They attribute this result to the plausible claims by managers of firms not reporting IC weaknesses

that they were unaware of the underlying conditions that led to misstatements, compared to cases

where the IC weakness had been previously disclosed. This suggests that the SOX 404 in operation

may not only permit firms to continue to maintain IC systems with weaknesses, but may even create

perverse incentives to hide IC weaknesses until a restatement forces the company to reveal them.

Far from being too stringent, section 404 as implemented may be milder than would be ideal for a

firm’s investors.

SOX-Like Laws in Other Countries

A third worry expressed by SOX skeptics was that the U.S. would drive public company

listings overseas, to regimes less burdened with regulation. Before we review data on that possible

effect in the next section, it is worth noting that many countries imitated the U.S. in adopting SOX-

like statutes or regulations following the market downturn in 2001. (Kim and Lu [2013] provide a

comprehensive listing of corporate governance reforms undertaken in 26 advanced and emerging

economies.) In 2006, for example, Japan adopted its own so-called ‘‘J-SOX’’ statute, the Financial

Instruments and Exchange Law, which contained provisions equivalent to those in both sections

302 and 404 of SOX. In 2006, the European Union, too, adopted an Eighth Directive on securities

disclosure, which largely tracked much of the contents of SOX. One nominal difference between

the way that the European directive was implemented in member states (such as the U.K. and The

Netherlands) was that officer certifications were required, as under SOX section 302, but audit firm

attestations as required in SOX section 404 were not.

More generally, a broader contrast is often asserted between the U.S. ‘‘mandatory’’ regime in

SOX and the ‘‘comply or explain’’ regime in the EU, and it is true that most of the EU member state

implementations include ‘‘comply or explain’’ components. But so too does the U.S. (as discussed

in the prior section). The stark contrast often suggested is an overstatement at best.

A more important distinction between the way that SOX was implemented in the U.S. and how

its equivalents were implemented elsewhere is that the ‘‘comply or explain’’ component is not well

enforced in many non-U.S. regimes. Van de Poel and Vanstraelen (2011) find, for example, that

while firms complying with the Dutch equivalent of SOX section 302 experience fewer abnormal or

SOX after Ten Years: A Multidisciplinary Review 635

Accounting Horizons September 2014

discretionary accruals in the period 2004–2005, non-complying firms’ accruals showed no

relationship with the extent that they explained why they did not comply, consistent with the

‘‘explain’’ aspects of Dutch law not generating market-relevant information or exerting a strong degree of pressure on Dutch firms. It would be interesting to apply the same method to U.S. firms

reporting persistent IC weaknesses to test whether the combination of U.S. disclosure requirements

and U.S. market forces are any more powerful than those in The Netherlands.

Why did other countries adopt SOX-like laws? In the absence of research evidence, one could

conjecture that, to an extent, the heightened awareness of corporate frauds revealed in the economic

downturn in early 2000s drove adoption of new laws to deter fraud. The governance template

provided by SOX made it easier for countries to copy the law. Global transmission of regulatory (or

best practice) guidelines has been observed before (e.g., board independence rules) and is an

interesting topic that has not received much attention in the accounting, finance or legal literatures.

SOX’S EFFECTS ON GOING PRIVATE, GOING DARK, GOING PUBLIC, AND CROSS- LISTINGS

Another set of predictions commonly advanced by critics of SOX is that it would increase the

marginal cost of being a U.S.-registered public company more than the benefits of that status,

causing existing public firms to go private or go dark, and deterring other companies from going

public or cross-listing in the U.S. A substantial number of studies have attempted to test these

predictions by looking at firm behavior before and after SOX’s passage. The going-dark,

going-private and cross-listing studies produce similar results—smaller, less liquid and more

fraud-prone firms did indeed exit U.S. stock markets after SOX—but the evidence that SOX

reduced the number of IPOs is weak at best, and is offset by evidence that IPO pricing improved.

More generally, the design of these studies is such that the relevance of their findings to an

assessment of SOX remains unclear, as changes in the number of public companies can only be

evaluated together with the propensity of those companies to commit fraud and the costs of such

fraud.

Going Private

Engel, Hayes, and Wang (2007) report that the propensity to go private spiked in the 33

months after SOX, relative to the same period prior to SOX (see Figure 2). This effect is

concentrated among smaller, less liquid firms with growth prospects that are lower than either prior

to SOX or as compared to companies that remain public post-SOX. They report a median market

value for firms going private post-SOX of $23 million, less than half than that of firms going private

pre-SOX, and less than one-seventh that of their overall CRSP/Compustat sample. Kamar, Karaca-

Mandic, and Talley (2009) confirm these findings in a difference-in-difference design, comparing

going-private deals by U.S. firms with those by foreign firms before and after SOX. They find no

increase for U.S. firms overall, but do find more going-private deals by U.S. firms with market

capitalizations under $30 million.

Interestingly, then, the majority of firms going private post-SOX were small enough that they

would never have had to comply with SOX section 404, which as noted above did not and

continues not to apply to $75 million or under firms. Indeed, Gao, Wu, and Zimmerman (2009) find

evidence that firms took steps to ‘‘manage’’ their market capitalizations to remain below this threshold, and so to remain exempt from section 404. This calls into question whether going-private

decisions immediately after SOX by small firms were in fact motivated by SOX. Such transactions

may instead have either been a response to contemporaneous changes in the financial and legal

environments (including low interest rates and a rise in private equity buyout activity), or a

mistaken expectation of how SOX would in fact be implemented over time, or both. Consistent

636 Coates and Srinivasan

Accounting Horizons September 2014

with this possibility is the fact that Leuz (2007) finds that going-private trends in the U.K. were

similar to those in the U.S. after SOX.

Bartlett (2006 ) further shows that going private does not per se exempt issuers from the Act, as

such transactions often entail issues of high-yield debt that require continued SEC reporting and

SOX-compliance. Empirically there was no post-SOX shift in buyout financing (e.g., toward bank

debt) that would have exempted newly private companies from the Act, except among firms under

about $220 million in book value. In fact, buyers of large firms were more reliant after SOX on debt

financing that triggers ongoing SOX-compliance even after the target firms have gone private.

Going Dark

Similarly, more companies chose to ‘‘go dark’’ (i.e., deregister their common stock and thus

suspend their SEC reporting obligations) after SOX (Leuz, Triantis, and Wang 2008), 370 from

2002–2004 versus 114 from 1998–2001 (see Figure 2). Leuz et al. (2008) are careful to exploit

several events in the phase-in and extended exemptions for small firms from section 404, making

their findings less likely to have been caused by contemporaneous changes in the legal and financial

environment. They find that firms that go dark are smaller, have poorer performance, weaker

growth opportunities, and are closer to financial distress, than firms that do not go dark. Leuz et al.

(2008) also find that that firms that go dark have weaker accounting quality, larger free cash flow

problems, and weaker board governance and outside monitoring. These findings raise the question

whether there is a net social cost or benefit from going dark (and similar) transactions, given such

FIGURE 2 Frequency of U.S. Firms Going Private and Going Dark

Going Dark data are from Leuz et al. (2008).

The figure shows the rate of U.S. firms going private (line with triangles) and going dark (line with squares). Going private numbers are from Engel et al. (2007).

SOX after Ten Years: A Multidisciplinary Review 637

Accounting Horizons September 2014

characteristics may indicate they pose more fraud risk, even if such transactions were privately

optimal for the managers of the firms involved and even if they were in fact caused by SOX. 5

Going Public

Some critics of SOX have also claimed that the costs of SOX, especially the burden of section

404 compliance on small firms, contributed to fewer IPOs in the U.S. in the 2000s (e.g., IPO Task

Force 2011). The drop-off in IPOs is most pronounced for small firms, consistent with a SOX

explanation. Bova, Minutti, Richardson, and Vyas (2013) document a related finding—U.S. private

companies, especially smaller ones, are more likely to be acquired rather than raise financing

through an IPO, post-SOX, unlike a control sample in the U.K., a result that could also contribute to

the IPO drop-off. 6

Such claims in the period leading up to 2012 motivated deferral of IC compliance for

‘‘emerging growth firms’’ in the JOBS Act. However, Gao, Ritter, and Zhu (2013) show that the

downward trend in IPOs started well before SOX (see Figure 3, where the drop in small firm IPOs

occurred in 2001, before SOX’s passage). Gao et al. (2013) also document that despite the SEC

revising rules to reduce the compliance costs burden on small companies, and then small firms

being permanently exempted by Congress in the Dodd-Frank Act, the number of small IPOs has not

increased. Gao et al. (2013) attribute the drop-off in IPOs to the absence of profitable small

companies and technological changes that make economics of scope and ability to speed products

to markets more important than in the past, giving an advantage to larger firms.

While the evidence of SOX’s effect on the number of IPOs in the U.S. remains contested, the

pricing of the IPOs that did occur after SOX improved. Johnson and Madura (2009) find that first-

FIGURE 3 Number of IPOs in the U.S.

The figure is from Gao et al. (2013).

The figure shows the frequency of IPOs in the U.S. by small (, $50 MM sales) and large firms, by year.

5 Leuz et al. (2008) show that once they separate the going-dark transactions from going-private transactions, there was no negative trend in going-private transactions. They suggest that the interpretation in Engel et al. (2007) arises from co-mingling going-dark and going-private transactions.

6 One can also observe a big spurt in IPOs by smaller firms in the mid to late 1990s. Given the higher rate of failure for small firms in general, one would expect a higher rate of de-listings by smaller firms. This poses a challenge to inferring a SOX effect on the higher rates of going-dark and going-private transactions post-SOX in the literature discussed earlier in this section.

638 Coates and Srinivasan

Accounting Horizons September 2014

day returns of IPOs in the U.S. have declined in the U.S. (but not in a matched sample of Canadian

IPOs not subject to SOX)—that is, U.S. IPOs have not been as underpriced as similar IPOs had

been in the past. In other words, IPO pricing became less uncertain, consistent with issuers

incurring lower capital costs as a result. They also find that one-year mean and median returns after

post-SOX IPOs ceased to be negative, as had characterized the pre-SOX period, which Loughran

and Ritter (1995) suggest was caused by a correction of first-day excess returns.

Cross-Listings

Concerns have also been expressed about fewer foreign firms choosing to list in the U.S. post-

SOX, a concern empirically validated in the data (see Figure 4). Piotroski and Srinivasan (2008)

find that fewer foreign firms enter U.S. equity markets post-SOX compared to pre-SOX, and that

defecting firms are more likely to list in London. Firms that defect to London are smaller, less

profitable, less likely to have a Big 4/5 auditor, and are more likely to be based in developed

countries. These firm characteristics are consistent with costs of SOX being higher for smaller

companies (as with U.S. firms going dark), while the benefits of a U.S. listing are lesser for firms

from developed countries.

However, the fewer numbers of listings hide the fact that the firms that choose to list in the U.S. post-SOX are larger, more profitable companies from less developed countries. Piotroski and

Srinivasan (2008) attribute this to the greater value from SOX-driven control and monitoring in the

U.S. for larger firms that increases the benefits of a U.S. listing, consistent with the bonding

hypothesis (Coffee 1999). In market capitalization terms, these larger cross-listing firms more than

FIGURE 4 Frequency of Foreign Firms Listing and Delisting From U.S. Exchanges.

Data from new listings are from Piotroski and Srinivasan (2008) and for delistings are from Hostak et al. (2013).

The figure shows the number of foreign firms listing in the U.S. over time (dark line) and the number of foreign firms delisting from the U.S. over time (light gray line).

SOX after Ten Years: A Multidisciplinary Review 639

Accounting Horizons September 2014

make up for the loss in listings by small firms, resulting in a net gain in market capitalization from foreign listings into the U.S., post-SOX. Bond market evidence is similar. Fewer foreign firms enter

U.S. public bond markets, but those that do also have equity listed in the U.S., are adopters of

International Financial Reporting Standards, and are larger bond issuers (Gao 2011).

Foreign firms are also materially more likely to delist from U.S. equity markets after SOX (see

Figure 4). Again, as with U.S. firms going dark, delisting firms are smaller and have low trading

volumes in U.S. exchanges (Marosi and Massoud 2008; Woo 2011). These firms had weaker

governance (e.g., less independent boards, higher separation of control and cash flow rights,

indicating greater agency problems), and were from countries with weaker investor protections

(Hostak, Karaoglu, Lys, and Yang 2013). While investors in these firms presumably suffered when

the firms deregistered, and managers and controlling shareholders benefited by cutting compliance

costs and market scrutiny, the social impact on the U.S. is less clear. Similar to domestic U.S. firms

that deregistered, foreign firms that delisted have characteristics that make them more likely to have

accounting problems and fraud.

SURVEY EVIDENCE—PERCEPTIONS OF SOX’S EFFECTS

Before we turn to ‘‘hard’’ evidence on the costs and benefits of SOX, we review ‘‘soft’’ evidence

from surveys of informed participants in the capital markets on the specific effects and the costs and

benefits of the law. A survey of 336 chief financial officers of companies considering ‘‘going public,’’

conducted in 2003 and published in The Journal of Finance by Brau and Fawcett (2006), showed that neither SOX section 404 nor SEC disclosure requirements more generally were a ‘‘significant

concern’’ for such officers, consistent with the findings of Gao et al. (2013) discussed above.

The SEC conducted a comprehensive survey on the economic effects of section 404 between

December 2008 and January 2009, which received responses from 2,901 companies representing

over half the companies filing Sec 404 (b) reports at that time. Based on this sample, Alexander,

Bauguess, Bernile, Lee, and Marietta-Westberg (2013) report a causal link between section 404

compliance and improvements in quality of the firms’ information environment, such as a positive

impact on IC structures, quality of financial reporting, and firm’s ability to prevent and detect fraud.

Despite respondents’ recognition of specific benefits, only 19 percent perceived a net benefit for

fiscal years completed after the 2007 reforms, with another 20 percent saying the costs and benefits

were equivalent.

The number perceiving a net benefit from section 404 rose to 26 percent for respondents with

fiscal years still in progress at the time of the survey, with another 20 percent believing the costs and

benefits were equivalent—suggesting a trend toward greater acceptance of section 404 over time.

Consistent with these findings and the existence of a learning curve under section 404, Alexander et

al. (2013) report that the number of years of section 404 (b) compliance, SEC’s management

guidance, and PCAOB AS5 are associated with perceived net benefits.

These findings are also consistent with findings in an interview-based study of corporate directors

reported in Cohen, Hayes, Krishnamoorthy, Monroe, and Wright (2013). The surveyed directors

report that SOX had a positive impact on the empowerment and authority of the audit committee and

in the responsibilities and status of internal auditors. Directors cautioned that the benefits were

tempered by a possible overreaction to SOX, driven by a perceived increase in litigation risk. They

also expressed concern that a compliance focus comes at the cost of appropriate risk management.

More broadly, a survey by the Financial Executives Research Foundation in 2005 (FERF

2005) found that 83 percent of large company CFOs agreed that SOX had increased investor

confidence, and 33 percent agreed it had reduced fraud. In a 2009 survey conducted by the SEC

(SEC 2009) of 2,907 firms, with a response rate of 35 percent—which is high for such a large

survey of businesses—27 percent agreed that SOX had enhanced investor confidence, but

640 Coates and Srinivasan

Accounting Horizons September 2014

interestingly, an even higher 40 percent agreed that SOX had made the respondents more confident

in other companies’ financial reporting.

In 2010, a Center for Audit Quality survey (n ¼ 1,001) (CAQ 2010) of investors reported that 65 percent were ‘‘concerned’’ about the permanent exemption of $75 million and under public

companies from SOX section 404, and in a 2012 Protiviti survey of public companies with more

than one full year of SEC filings, 50 percent believed that SOX’s benefits outweighed or were

equivalent to its costs, 47 percent did not think that small public companies should be exempt from

SOX section 404(b), and 62 percent did not believe section 404(b) should be eliminated for

companies under $1 billion in market capitalization (Protiviti 2012).

Finally, the GAO (2013) surveyed 746 public companies with a response rate of 25 percent,

specifically focused on SOX section 404. On the one hand, the GAO reported that 80 percent of all

companies viewed auditor attestation under section 404(b) as ‘‘benefiting’’ the quality of the

company’s controls, 53 percent viewed the requirement as benefiting their company’s financial

reporting, 46 percent viewed their ability to prevent and detect fraud as benefiting from section

404(b), and 52 percent reported greater confidence in the financial reports of other section 404(b)-compliant companies. On the other hand, only 30 percent reported that section 404(b) raised

investor confidence in their own company, and only 16 percent believed that section 404(b)

increased their company’s ability to raise capital.

In sum, contrary to vehement criticism of SOX in some media reports and analyses by

political entrepreneurs (and politically active academics), the reception of SOX among the

constituencies most affected by SOX has been far more nuanced, even receptive, consistent with

Langevoort (2007)’s speculation that market and other incentives would blunt managerial

hostility to the Act.

EVIDENCE ON SOX’S COSTS

We turn now to ‘‘hard’’ evidence on the costs of SOX. These costs include direct costs, such as

expenditures on IC systems or litigation, and indirect costs, such as those that may have arisen from

SOX’s effects on risk-taking, investments, and other activities. We consider which elements of

SOX have contributed to these costs, and assess how well research has been able to attribute the

observed effects directly to SOX, as opposed to other changes in the legal, accounting, regulatory,

and market environment for firms that occurred close in time to SOX.

Direct Costs from Control System Expenditures

The direct costs include expenses for IC testing and reporting and audit fees to attest to IC

effectiveness. Such costs clearly increased as a result of SOX, but how much remains unclear. The

SEC had estimated a cost of $91,000 per filer for internal section 404(a) compliance, but did not

have a basis for estimating the costs of auditor attestation under section 404(b) (SEC 2004). 7

7 Some have criticized the SEC for supposedly grossly underestimating the costs of SOX section 404. Hal S. Scott, Director, Committee on Capital Markets Regulation, in his statement to the U.S. House of Representatives Small Business Committee stated that the SEC’s estimate is now known to be have been off by a factor of 48 (Scott 2007). Similarly, Representative Patrick T. McHenry stated in hearings before the Subcommittee on Regulatory Affairs of the U.S. House of Representatives that the SEC’s estimates were off by a factor of 40. (Committee on Government Reform 2006 ). However, these criticisms mis-describe the SEC’s analysis, which estimated the internal cost of producing a control report satisfying section 404(a), and not the external cost of obtaining an attestation from a company’s auditors on a company’s control report, as required by section 404(b) (SEC 2004 at V.B. and note 174 [‘‘The estimate does not include the costs of the auditor’s attestation report, which many commenters have suggested might be substantial’’].) On the latter cost, the SEC simply noted that it did not have adequate information to publish an informed estimate.

SOX after Ten Years: A Multidisciplinary Review 641

Accounting Horizons September 2014

Estimates of direct costs of SOX are largely derived from surveys, but these surveys do not use the

same methods or samples, and rely on subjective and potentially biased assessments by respondents

on items such as time spent by management overseeing compliance changes. Despite their

uncertainties, such surveys do indicate that SOX’s costs were and remain increasing in firm size,

but at a decreasing rate, so that larger firms pay less for SOX-related services as a percentage of firm

size than do smaller firms (Coates 2007; Cox 2013).

Direct costs have also been steadily falling over time for all firms. Charles River Associates

(CRA) estimated, for example, that Fortune 1000 firms spent $5.9 million, on average, to comply with section 404 in the first year of implementation (CRA 2005). By the second year, however,

similar surveys estimated total direct costs had fallen by between 15 percent and 40 percent and

varied significantly by size of firm (see Coates 2007, 107). Subsequent surveys show further drops

of about 25 percent following the 2007 revisions to section 404’s implementation by the PCAOB,

with the result that by 2012, firms with less than $10 billion of market capitalization were paying an

average of $350,000 for section 404 attestation, representing about 30 percent of total audit fees

(SEC 2009; GAO 2013). The SEC survey sample in Alexander et al. (2013) showed an overall

average section 404 compliance expense of $1.2 million in the latest fiscal year before the survey

(December 2008–January 2009) and respondents reported a decline over time in such costs.

For at least four reasons, comparisons of audit fees and/or internal costs before and shortly after

SOX are likely to exaggerate SOX’s actual long-term direct costs. First, they fail to take into

account the fact that audit fees were rising prior to SOX (Asthana, Balsam, and Kim 2009), and

would likely have continued to rise as a result of the financial failures that led to SOX, so that some

costs attributed to section 404 would have been incurred without SOX. Second, they fail to reflect

the fact that firms were already required prior to SOX to have effective IC systems, as noted above,

but many did not, so some costs attributed to SOX should instead be attributed to firms ‘‘catching

up’’ on their legal obligations in an environment of increased enforcement. Third, in the immediate

post-SOX period, companies and audit firms were learning rapidly how to better assess and report

on their IC systems, resulting in start-up costs and fixed investments with a long potential payoff

period being wrongly identified as recurring annual costs. Fourth, they do not reflect the significant

cost-saving revisions to the implementation of SOX section 404 (b) by PCAOB in 2007, discussed

above.

To partly address the first two of these issues, Iliev (2010) provides a nicely designed empirical

estimate of direct costs of section 404 implementation using a regression discontinuity design,

examining the increase post-SOX in audit fees above and below the ‘‘small firm’’ exemption from

section 404 for firms with a market capitalization of $75 million or less. He finds that, for small

firms just above or below the $75 million cut-off, filing a control effectiveness report under section

404(a) in 2004 increased audit fees (due to the attestation requirement in section 404 [b]) by 86

percent, or roughly $528,000 for the mean firm with a market capitalization of about $135 million.

However, as reflected in Figure 5, his findings by design cannot be generalized to the larger firms

that are the majority of companies subject to SOX and on which greater SEC scrutiny was brought

post-SOX (as discussed below). This creates a problem of external validity since SOX costs fall and

benefits likely increase as a percentage of size. In addition, his research design does not address the

second of the two issues reviewed above. Survey results from more recent periods are likely a better

approximation of long run, equilibrium direct costs of SOX, and, as noted above, on average they

are much lower than Iliev’s (2010) small-firm estimate.

Direct Costs from Securities Litigation Related to SOX

Another type of direct cost predicted to result from SOX is litigation-related costs. Lawsuits

brought under U.S. securities laws might have become easier to pursue (or use to extract

642 Coates and Srinivasan

Accounting Horizons September 2014

settlements) as a result of SOX sections 302 and 404, which together force more disclosures that

may be used by plaintiffs and their lawyers to identify defendants and to present in court as

evidence of fraud. Critics expected that SOX would generate more litigation—indeed, some went

so far as to predict an explosion of litigation as a result of SOX, which they called a ‘‘litigation time bomb’’ (Butler and Ribstein 2006 ). There was also concern about higher litigation risk for independent directors due to greater responsibilities under SOX, especially for audit committee

directors (Bebchuk et al. 2006 ).

Securities litigation incidence did rise from 2002 to 2004 following SOX (see Figure 6, Panel

A). However, it dropped to pre-SOX levels soon thereafter, with no subsequent change in law that

could account for the drop. Similarly, litigation risk for independent directors spiked in 2002, but

then reverted to pre-SOX levels. Brochet and Srinivasan (2013) examine litigation risk for

independent directors and document no increase over time in such risk (see Figure 6, Panel A).

Settlement amounts also increased in the period 2004–2006 (see Figure 6, Panel B). Likely

reflecting this proximate trend, director and officer (D&O) insurance premiums increased

substantially. Linck, Netter, and Yang (2009) found that median D&O insurance premiums

increased by more than 150 percent from 2001 to 2004. However, this evidence is restricted to a

small sample of firms, since D&O insurance is only required to be reported for firms incorporated in

New York state. However, with no change in the underlying law, securities litigation settlements

began to plummet in 2007, and have remained steady at relatively low, pre-SOX levels since 2008.

This pattern is not consistent with SOX creating a powerful array of litigation tools for shareholders

to sue companies or their managers, or for plaintiffs’ lawyers to use on behalf of shareholders in

class actions.

Instead, the pattern is more consistent with large costs arising from the fundamental corporate

misconduct that gave rise to SOX, followed by a reduction in that misconduct (at least as perceived

by litigants and other participants in the legal system). This pattern holds even after taking account

FIGURE 5 Share of Public Firm Market Capitalization in Iliev (2010)

SOX after Ten Years: A Multidisciplinary Review 643

Accounting Horizons September 2014

of three of the largest all-time settlements (Enron, WorldCom, and Tyco). Whatever the costs of

SOX, increased litigation does not appear to be one of them.

Indirect Costs of SOX

In addition to (and likely more important than) the direct costs discussed above, SOX also

imposed indirect costs on firms. While direct costs can be significant, especially for small firms,

indirect costs can be large for all firms if managers and boards of directors lowered investments and

risk-taking due to fear of the increase in director and officer liability and/or due to distraction from

core business concerns by the increased focus on internal controls. Alan Greenspan noted in 2003

that ‘‘business leaders have been quite circumspect about embarking on major new investment

projects’’ (Greenspan 2003). Echoing similar concerns, William Donaldson, former chairman of the

SEC predicted that SOX would lead to a ‘‘loss of risk-taking zeal’’ and has created a ‘‘huge

preoccupation with the dangers and risks of making the slightest mistake’’ (Michaels 2003).

Some studies have attempted to back up these claims with detailed findings. Kang, Liu, and

Qi (2010) find that the investment to capital ratio declined for U.S. firms compared to a sample

of U.K. firms after 2002. This evidence is corroborated by Bargeron, Lehn, and Zutter (2010),

FIGURE 6 Securities Litigation Before and After SOX

Panel A: Litigation Rate for U.S. Firms (Solid Line) and Litigation Rate for Independent Directors on the Board of U.S. Firms (Dotted Line)

Data for the graph is from Brochet and Srinivasan (2013).

(continued on next page)

644 Coates and Srinivasan

Accounting Horizons September 2014

who find several measures of risk-taking decline for U.S. firms compared to a sample of publicly

listed U.K. and Canadian firms over the period 1994 through 2006. Similar to the trends

observed in Figure 7 (from Albuquerque and Zhu 2013), they find that U.S. firms reduced their

investment in capital and research and development (R&D) expenditures, increased their cash

holdings, and reduced leverage over this time period. Standard deviation of stock returns for

U.S. firms also fell compared to non-U.S. firms. These changes are greater for larger firms

compared to smaller ones.

However, as with the decline in U.S. IPOs depicted above, Figure 7 shows that the trends in

investments and cash holdings manifested before SOX. In fact, Bates, Kahle, and Stulz (2009)

document a trend of increasing levels of cash holdings and lower leverage for U.S. firms starting in

the early 1980s. Albuquerque and Zhu (2013) contest the findings of SOX-driven lower

investments using a regression discontinuity design similar to Iliev (2010). They compare firms that

were just above and below the threshold of $75 million in market capitalization for section 404

implementation. They find that investment declines post-SOX for small firms subject to section

404. However, they find a similar decline in investment for the companies not subject to section

404, suggesting an overall decline in investment in U.S. companies. As shown in Figure 7, they find

that the decline begins in 1999 or even earlier and not in 2003, when SOX went into effect.

In other words, comparisons with the U.K. and Canada in Kang et al. (2010) and Bargeron et

al. (2010) likely capture a trend of declining investment in the U.S., but one that is unrelated to

FIGURE 6 (continued)

Panel B: The Figure Shows Total Settlement Amounts in Securities Class Actions between 2003 and 2012 in $ Million

The figure is from Ryan and Simmons (2012).

SOX after Ten Years: A Multidisciplinary Review 645

Accounting Horizons September 2014

SOX. Indirect costs remain possible side effects of SOX, but their size and significance remain

unclear.

EVIDENCE ON SOX’S BENEFITS

Next, we survey the evidence on the benefits of SOX.

Impact on Accounting Quality

A number of papers provide evidence consistent with the observation that accounting quality

improved for U.S. firms in the post-2002 time period. Cohen, Dey, and Lys (2008) find that accrual-

based earnings management increases steadily from 1987 until 2002, then declined significantly.

Some part of this decline has been offset by an increase in real earnings management (as measured

FIGURE 7 Risk Taking and Investments

Panel A: Capital Expenditure Scaled By Total Assets

Panel B: R&D Expenditure Scaled By Total Assets

(continued on next page)

646 Coates and Srinivasan

Accounting Horizons September 2014

by abnormal cash from operation, abnormal production costs, and abnormal discretionary

expenses), which had been declining prior to SOX. Likewise, firms that just met earnings

benchmarks used less accrual and more real earnings management post-SOX compared to similar

firms before SOX (Chhaochharia and Grinstein 2007; Lobo and Zhou 2006 ).

Despite the offset in higher real earnings management, other studies suggest a general

improvement in accounting quality. Koh, Matsumoto, and Rajgopal (2008) and Bartov and Cohen

(2009) find that firms exhibited a lower tendency to engage in the phenomenon of just meeting or

beating analysts’ consensus forecasts by managing earnings. Lobo and Zhou (2006 ) document a

significant increase in timely loss recognition, a measure of how quickly accounting incorporates

economic losses. Daniel, Denis, and Naveen (2008) find that firms manage earnings less for

dividend payout reasons. Dyck, Morse, and Zingales (2010) find that auditors greatly increased

their role in detecting and reporting fraud in public companies after SOX, relative to other sources

FIGURE 7 (continued)

Panel C: Cash Holdings Scaled By Total Assets

Panel D: Standard Deviation of Returns

The figure is from Albuquerque and Zhu (2013).

SOX after Ten Years: A Multidisciplinary Review 647

Accounting Horizons September 2014

of fraud detection, from 6 percent of total frauds to 24 percent in the post-SOX period. This increase

is spread across not only restatement cases, but also those not involving restatements. Kedia, Koh,

and Rajgopal (2013) find that prior to 2002, companies were more likely to copy poor accounting

practices (that lead to a restatement) upon observing restatements by industry peers. This behavior

appears to have stopped after 2002, which the authors conjecture is driven by more rigorous

enforcement after SOX.

Further, some evidence suggests these improvements may be driven by a combination of SOX

and changing behavior in stock markets and inside boardrooms. Brochet (2010) finds that insider-

trading disclosures became more informative after SOX. CEO and CFO certifications seem to have

helped as well. Qian, Strahan, and Zhu (2009) find that firms with public floats above the trigger for

SOX section 404 ($75 million), but below $350 million, reduced CEO compensation but increased

insider ownership, while firms below $75 million increased CEO pay and did not increase insider

ownership in the post-SOX period. Stock market premium to meeting analysts’ consensus forecasts

disappeared post-2002 according to Koh et al. (2008). In fact, post-2002, stock prices react

negatively to positive earnings surprises by firms the CEOs of which have higher equity incentives.

Similarly, Jiang, Petroni, and Wang (2010) find that CEO equity incentives, considered a culprit in

providing perverse earnings management incentives to managers to artificially boost stock prices,

are not related to proxies for earnings management post-SOX in contrast to the pre-SOX time

period. They find that neither CEO nor CFO equity incentives are positively related to magnitude of

accruals during the 2002–2006 period in contrast to the prior period—1993–2001. In fact, the

relationship of accruals to CFO incentives reverses—higher incentives are associated with a lower

magnitude of accruals. Boards seem to take the apparently improved managerial behavior into

account when setting pay-sensitivity of bonus pay to accounting performance becomes higher, as

documented in Carter, Lynch, Zechman (2009). They also find that firms place more weight on

positive earnings changes post-SOX, and that firms with the largest decrease in positive

discretionary accruals have the largest increase in weight on earnings changes. Significant increase

in sensitivity of CEO cash compensation to accounting performance suggests that boards trust

accounting numbers more.

One of the precursors to SOX was the steady rise in restatements prior to 2002. After SOX, the

number of restatements initially increased dramatically and the increase is mostly attributed to the

increased vigilance post-SOX. The rate of restatements fell as dramatically as they rose (Figure 8,

Panel A). On average, restatements have been less egregious after SOX, as seen in lower dollar

amounts, more unintentional errors, more noncore accounts, and a lower average negative market

reaction to restatement announcements after controlling for the lesser intensity (Burks 2010, 2011;

Hennes, Leone, and Miller 2008; Plumlee and Lombardi Yohn 2010; Scholz 2008). Finally, the

extent of adverse SOX section 404 auditor attestation opinions (as a fraction of all companies to

which section 404 applies) has been declining over time, ranging from a high of 16.9 percent in

2005, when the mandates first kicked in, to 2.4 percent in 2010 (see Figure 8, Panel B).

The main weakness of these streams of research on SOX’s benefits (as with studies of SOX’s

costs, discussed above) is that, while the effects documented are for the time period following SOX,

the accounting quality benefits found in this literature may not have been caused by SOX itself. Critics can validly claim that several benefits could have arisen from market discipline following a

period where internal controls were widely seen to have collapsed. Research designed to isolate the

effects of SOX is challenging since the Act affected all large SEC registered firms in the U.S.,

depriving researchers of a control group of firms that was unaffected by SOX and yet comparable to

the majority of U.S. public companies. Several contemporaneous changes, such as changes in stock

exchange guidelines, provide alternate explanations for the observed effects. Finally, SOX has

several complementary provisions making it difficult to identify the provisions that contributed to

the changes in observed behavior.

648 Coates and Srinivasan

Accounting Horizons September 2014

Impact of Section 404 Disclosures

Other research more directly documents the benefits of SOX using information in SOX-driven

disclosures. IC weakness reports mandated by SOX have proved informative to investors. Firms

with IC weaknesses have accruals that do not map well into cash flows, more auditor resignations,

and more restatements and SEC enforcement actions; they also provide less precise management

forecasts, and have CFOs with weaker qualifications (Doyle, Ge, and McVay 2007; Ashbaugh-

Skaife , Collins, Kinney, and LaFond 2008; Feng, Li, and McVay 2009; Li, Sun, and Ettredge

2010). Not surprisingly, investors react negatively to disclosure of IC weakness reports increasing

both the cost of debt (Kim, Song, and Zhang 2011) and of equity (Hammersley et al. 2008).

Lenders are seen to modify debt contracts to rely less on financial numbers in covenants when firms

disclose IC weaknesses and instead substitute them with price, security, and credit-rating-based

protection (Costello and Wittenberg-Moerman 2011). Because such studies link changes in

behavior to specific requirements in the Act, they more plausibly reflect causation, rather than

temporal correlation, but they also leave a greater gap between the results and a quantitative

assessment of the value of the benefits.

FIGURE 8 Restatements and SOX 404 Adverse Reports

Panel A: Time-Series of Reported Restatements

Data are from Audit Analytics; Srinivasan, Wahid, and Yu (2014); and Cheng, Srinivasan, and Yu (2013). Sued restatements represent restatements that are followed by a securities class action lawsuit, thereby indicating a greater severity of the restatement compared to restatements that are not followed by a lawsuit. All restatements include sued and non-sued restatements.

(continued on next page)

SOX after Ten Years: A Multidisciplinary Review 649

Accounting Horizons September 2014

While the IC disclosure mandate in section 404 has come in for the most criticism for the costs

imposed on companies, better controls have been found to be helpful in improving accounting

quality. Using a sample of banks unrelated to SOX, Altamuro and Beatty (2010) find that rules

adopted in 1992 requiring internal control reporting for banks led to greater validity of loan loss

provisions in banks. The association between loan-loss provisions and actual loans written off

strengthened after the law was passed, and IC reporting led to greater earnings persistence, better

ability of earnings to predict cash flows, and a reduction in use of earnings management to report

positive earnings growth but also to lower earnings conservatism. They conclude that banks

exercised less reporting discretion after IC reporting was introduced. Not surprisingly, Ashbaugh-

Skaife et al. (2008) find that earnings’ properties improve after remediation of ineffective internal

controls. Remediation involves changes to management and boards, and leads to hiring of better-

qualified CFOs (Li et al. 2010; Johnstone et al. 2011). Firms also improve their investment

efficiency after remediation of IC weaknesses (Cheng, Dhaliwal, and Zhang 2013).

The implementation schedule of section 404 reporting allowed for studies that can produce

more compelling evidence of causal effects, at least over certain subsets of firms. Evidence from

such studies reinforces the view that section 404 causally improved accounting quality. Iliev (2010)

finds that section 404 filers just above the $75 million exemption from section 404 had significantly

lower accruals and discretionary accruals in 2004 compared to firms just below the implementation

threshold, which he interprets as more conservative reporting. Arping and Sautner (2013) find a

FIGURE 8 (continued)

Panel B: Percentage of Firms Receiving Adverse Sec 404 Auditor Attestations of Internal Control Weaknesses as a Proportion of All Companies

Data are from Audit Analytics.

650 Coates and Srinivasan

Accounting Horizons September 2014

reduction in analyst forecast error and forecast dispersion for European firms listed in the U.S. and

subject to SOX more than for matched foreign firms not cross-listed in the U.S.. These results are

driven by the time period when foreign firms were required to comply with provisions of section

404 (after 2006 and not before).

Impact of SOX on Audit Quality

Accounting quality likely also improved because of improvements in audit quality. DeFond and

Lennox (2011) find that, post-SOX, a number of small audit firms exited the market for public

company audits. Roughly half (607) of 1,233 small audit firms (small firms being defined as those

with fewer than 100 SEC clients) exited the public audit market following SOX. Small auditors

comprise 97 percent of all audit firms and audit 34 percent of public companies. Clients move to other

smaller auditors—with remaining small auditors doubling their average client base—but compared to

the non-exiting auditors, the exiting auditors are lower quality. Clients of exiting auditors receive

higher quality auditing from successor auditors, and successor auditors are more likely than exiting

auditors to issue going-concern opinions to the new clients. The authors conclude that PCAOB

inspections under SOX improve audit quality by incentivizing low quality auditors to exit the market.

Raghunandan and Rama (2006) document a significant increase (74 to 86 percent) in audit fees after

SOX, consistent with both a rise in audit effort and increased work load attributable to SOX 404

implementation. Griffin and Lont (2007) show that this increase reflected higher audit risk, suggesting

that SOX increased the auditors’ share of risk of defective financial statements. (See also DeFond and

Zhang [2013] for a review of audit research that focuses on SOX.) PCAOB inspections appear to

have played a role in improving audit quality. DeFond and Lennox (2011) attribute the exit of smaller

and lower quality audit firms to the threat of PCAOB inspections. Abbott, Gunny, and Zhang (2013)

find useful information content in PCAOB inspections—smaller auditors that receive adverse

PCAOB inspection reports have clients with accounting restatements and poor accrual quality.

However, inspection reports are not found to lead to auditor switching by clients, suggesting that

clients do not find the inspection information useful in selecting auditors (Lennox and Pittman 2010).

PCAOB sanctions against auditors are seen to be informative to investors and prompt audit firms to

improve audit quality. Dee, Lulseged, and Zhang (2011) examine PCAOB’s 2007 sanction of

Deloitte for an audit failure in 2004 that revealed quality control problems at Deloitte’s San Diego

office as well as nationally. They find that Deloitte clients suffered a negative 0.83 percent three-day

return surrounding the announcement of the sanction, which the authors attribute to the quality control

problems identified at Deloitte; following the sanction, the audit firm undertook governance changes

to strengthen oversight of its partners and directors.

EVENT STUDIES OF NET SHAREHOLDER WEALTH EFFECTS

The above combination of evidence on costs and benefits of SOX is hard to bring down to a

single bottom line. Some of the direct costs can be reduced to dollars, but there is substantial

variation across firms and over time even in those costs. Other direct costs, and all of the benefits

and indirect costs, are not easily translated into bottom-line dollar costs. As a result, the costs and

benefits cannot be totaled into a single net effect.

Partly because of these limitations, some scholars have attempted to use a single method to

estimate both costs and benefits. These studies have looked to stock market reactions to critical events

in the legislative process leading up to the enactment of SOX and test market reaction around these

events. Unfortunately, these studies have proven no more useful in resolving the policy assessment of

SOX, producing widely disparate results and remaining subject to significant uncertainties.

On the one hand, Zhang (2007) finds that stock prices decline in response to events leading up

to the passage of SOX. Firms with weaker shareholder rights, higher pre-SOX non-audit fees, more

SOX after Ten Years: A Multidisciplinary Review 651

Accounting Horizons September 2014

extensive foreign operations and larger abnormal accruals experienced greater negative returns. On the

other hand, Jain and Rezaee (2006), Li, Pincus, and Rego (2008), and Akhigbe and Martin (2006)

document that stock prices react favorably to the events leading up to the passage of SOX. Stock prices

react favorably for firms with higher levels of pre-SOX governance, for those with more reliable

accounting and higher quality auditing. There is a positive relationship of returns with extent of earnings

management and a positive stock price impact for financial services firms except for securities firms.

Again, the staggered implementation of section 404 provides the best opportunity to find an

effect most directly attributable to SOX. Iliev (2010) finds that non-accelerated filers experience

positive returns when the section 404 compliance deadline was extended. This implies that

investors respond as if the costs outweigh the benefits. While these results are able to capture the net

effect of the law, the findings are limited to relatively small firms that qualified for the extended

compliance deadline, and are not valid as an estimate of the overall effect of SOX on public firms

generally. In addition, this paper and the other event studies noted above use different sets of events

and present different and somewhat subjective interpretations of whether given events improved or

reduced the odds of SOX’s passage. 8

Leuz (2007) identifies a number of non-SOX related negative news events on the dates

identified in Zhang (2007) (e.g., declaration of Iraq war) that make it hard to identify SOX as the

cause of the market reactions studied. More generally, event study methods are better suited for

clearly defined events and not events that depend upon the opaque process of legislation. Finally,

implementation of SOX took place over time and, as described above, involved several decisions

that were not easily foreseeable at the time of SOX’s adoption—including PCAOB’s 2007 changes

in section 404’s implementation and the deferral and eventual exemption of small issuers altogether

from section 404—making it difficult for the market to assess how the changes would unfold.

OTHER RESEARCH RELEVANT TO EVALUATING SOX

Ban on Consulting Services by Audit Firms

Accounting literature has consistently diverged from the views of policy-makers and found scant

evidence that non-audit fees adversely affect financial reporting quality. For example, DeFond,

Raghunandan, and Subramanyam (2002); Chung and Kallapur (2003); Ashbaugh-Skaife, LaFond,

and Mayhew (2003); Kinney, Palmrose, and Scholz (2004); and Koh, Rajgopal, and Srinivasan

(2013) all find no effect of non-audit fees on financial reporting quality. Indeed, by banning such

services, SOX has made further study difficult due to the elimination of observed variation in practice.

Role of SEC Enforcement

SOX expanded SEC resources and gave the SEC more legal tools, but the political

environment as it relates to fraud shifted concurrently with SOX, making it possible that SEC

priorities would have shifted without regard to SOX. 9

Descriptive evidence suggests that the SEC

brought more accounting and auditing enforcement cases and pursued larger companies in the

8 Some papers provide event study evidence on the CEO/CFO certification requirement in Section 302 of SOX, again with mixed evidence. Griffin and Lont (2005) find that the certification event is value relevant whereas Bhattacharya, Groznik, and Haslem (2007) find that the certification was a non-event for certifiers around their certification date. Again, differing methodologies in these papers do not allow for reconciliation of these results.

9 There is limited research on the role of SEC enforcement on improving accounting quality in companies. One notable exception is Kedia and Rajgopal (2011). They find that firms located closer to SEC offices and in areas with higher prior SEC enforcement activity are less likely to restate their financials. Moreover, the SEC is more likely to investigate firms located closer to its offices. However, Jennings, Kedia, and Rajgopal (2013) find that the deterrence effect of SEC enforcement actions is not reflected in improved earnings quality at peer firms as measured by abnormal accruals, earnings response coefficients, or timely loss recognition.

652 Coates and Srinivasan

Accounting Horizons September 2014

aftermath of the scandals in 2002, which was no doubt supported by the increased budget (see

Table 1). The median size of a company that is the subject of an accounting and auditing

enforcement release (AAER) shows a clear upward movement from the pre-2002 to the post-2004

time period. Note that Table 1 presents data as of the AAER issuance date, which is typically well

after the fraud occurred. We observe a similar pattern (untabulated) when we consider when the

fraud actually occurred.

Consistent with these descriptive data, Cox and Thomas (2005) find that the SEC appears to be

targeting much larger companies in the post-2002 period than it did from 1990–2002. They

document that the average (median) market capitalization for SEC enforcement targets was 23 (13)

times bigger post-2002. Dyck et al. (2010) find that the SEC became more likely to be the first

detector of corporate frauds after SOX, increasing its share of total fraud detection (relative to other

actors) from 5 percent to 10 percent.

The pattern observed in Table 1 and by Cox and Thomas (2005) could occur if the type of

firms committing fraud has changed over time, with larger firms being more prone to commit fraud

now than before. Anginer, Narayanan, Schipani, and Seyhun (2012) address this issue by

examining firms that presented evidence of options backdating based on suspicious timing of option

grants on dates where the stock price was at its lowest. Based on this sample of all potential

backdating cases, they find that smaller firms were more likely to indulge in the practice than larger

firms. However, they find that larger firms were more likely to be prosecuted by the SEC and DOJ.

The effect of changing enforcement priorities on propensity to commit fraud in particular, and on

accounting quality in general, is an open issue and remains an unexplored potential consequence of

(and/or confound for) SOX. For example, Section 408 of SOX required the SEC to review periodic

financial statements of companies at least once every three years, representing a major step up from

then existing review practice. The resulting publicly available comment letters present a rich source

of data on financial reporting issues and a potential area for research.

TABLE 1

SEC Enforcement Actions

Year Number of AAERs Median Asset $ MM

1995 35 32.3

1996 40 16.8

1997 41 36.8

1998 17 64.2

1999 32 73.8

2000 44 79.0

2001 28 101.4

2002 60 163.9

2003 55 143.0

2004 59 526.3

2005 44 785.0

2006 47 717.2

2007 46 1967.2

2008 27 1833.8

2009 47 543.1

Data used for this table is the same as in Dechow, Ge, Larson, and Sloan (2011) and were generously provided by Patricia Dechow. The table provides descriptive statistics on accounting and auditing enforcement actions (AAER) by the SEC over time. Median Asset refers to the median total asset of the firms for which an AAER was issued in that particular year.

SOX after Ten Years: A Multidisciplinary Review 653

Accounting Horizons September 2014

Role of Boards and Audit Committees

Section 301 of SOX required that the audit committee be comprised exclusively of independent

directors and charged the audit committee with direct responsibility for the appointment,

compensation, and oversight of the auditor. There is a preponderance of evidence (from mainly

association studies) that more independent audit committees demand better audit quality and are

associated with better financial reporting quality (e.g., Klein 2002; Abbott, Parker, and Peters

2004). The previously discussed evidence on improvement in accounting quality post-SOX is

consistent with improved monitoring by audit committees, though direct evidence of SOX

mandates on audit committee independence contributing to this effect is lacking.

Section 407 of SOX required firms to comply with the requirement to have at least one audit

committee financial expert, or explain why they did not, leading to the question of the role financial

experts play in improving financial reporting. DeFond, Hann, and Hu (2005) examine 3-day market

reaction around director appointments to the board pre-SOX (1993–2002) and find a positive

reaction to appointment announcement for accounting financial experts, but no reaction to non-

accounting financial experts or non-financial directors, suggesting that the market places a positive

value on accounting experts. They document a positive reaction even if the board had a prior

financial expert. The paper finds weak evidence that the market responds to expectations of better

monitoring than the signal of quality when an expert joins the board. Accounting experts on the

audit committee, in particular, improve accounting quality. Dhaliwal, Naiker, and Navissi (2010)

find that audit committee accounting expertise is positively associated with accruals quality for

audit committee accounting experts with lower tenure and fewer other directorships. Several papers

use board changes prompted by changes in stock exchange listing requirements to examine the

impact of board characteristics on a variety of firm outcomes. While interesting, we do not discuss

these papers, as they do not directly relate to the effect of SOX itself, which changed the

requirements only for audit committees.

Effect of SOX on Governance and Transparency Generally

Hochberg, Sapienza, and Vissing-Jorgensen (2009) use an innovative approach to study wealth

effects by examining stock market reaction for firms that lobby for or against the provisions of

SOX. Greater returns to firms that lobby against strict implementation of SOX suggest SOX had a

positive impact on corporate transparency and governance. Further, they find that firms with

corporate insiders that oppose strict SOX-related rules are larger, more profitable, have lower future

growth opportunities, and retain more cash, all characteristics of free cash flow or agency problems.

These firms experience higher cumulative abnormal returns during the SOX passage period relative

to peer firms that choose not to lobby. These cumulative returns do not reverse themselves

following the passage of SOX.

Auditor Performance in the Financial Crisis

A last set of studies bears on how well auditors performed in the financial crisis as one

measure of whether SOX accomplished the goal of improving accounting and auditing. By that

measure, the results are mixed. On the one hand, within a short time of receiving clean audit

reports (and section 404 control attestations) from their auditors, many financial institutions

failed in contexts calling into question their prior accounting. As reviewed in Sikka (2009),

Bear Stearns received a clean report from Deloitte on January 28, and failed on March 14;

Thornburg received a clean report from KPMG on February 27, which was withdrawn on

March 4, prior to Thornburg’s failure; Northern Rock received a clean report on July 25, only

to suffer a bank run on September 14; and U.S. Bancorp received a clean report from E&Y on

654 Coates and Srinivasan

Accounting Horizons September 2014

February 20, before receiving $6 billion in bailout funds on November 14. As detailed by a

bankruptcy court examiner (Valukas 2010), E&Y ‘‘were aware of but did not question Lehman’s use and nondisclosure’’ of ‘‘window-dressing’’ repo transactions to ‘‘manage’’ its balance sheet around period ends,

10 to make it appear that it was smaller and less leveraged

than it was, leading the examiner to conclude there was ‘‘sufficient credible evidence to support a finding by a trier of fact’’ that E&Y had failed to meet professional auditing standards. Consistent with these examples, Desai, Rajgopal, and Yu (2013) find large sample evidence

that auditors showed no ability to assess bank risk and predict weaknesses in the lead up to the

financial crisis in contrast to the ability of short sellers to do the same.

On the other hand, a few studies suggest that some of the goals of SOX improved financial

institutions in the crisis. Yeh, Chung, and Liu (2011) find in a cross-country study that banks with

fully independent audit committees prior to the crisis performed better, measured by total return to

shareholders and return on equity during both 2007 and 2008. Jin, Kanagaretnam, and Lobo (2011)

found that, among small and mid-sized U.S. banks, those with auditors that specialized in banking

in 2006 were less likely to fail in the crisis (i.e., 2007 to 2009). Doogar, Rowe, and Sivadasan

(2013) show that audit firms increased fees charged to their banking clients to reflect the changing

risk profile of their loan portfolios and securitization risk over the time period from 2005 to 2007,

suggesting that audit firms recognized client risk and priced their services accordingly. The authors

attribute the failure by audit firms to provide advance warning of client failure to limitations in the

auditor’s report and the scope of the auditors’ disclosure requirements rather than to a failure to

assess client risk. Together, the evidence suggests that better auditing and more independent

auditors did help alleviate the worst effects of the crisis, even if SOX, PCAOB, and audit firms

performed well below the ideal.

A RESEARCH AGENDA: SOX AND BEYOND

Drawing on the prior sections, we develop a research agenda that better grapples with

difficulties of measuring the full set of costs and benefits caused by and not merely correlated with

new financial regulations such as SOX. To accomplish this general goal, multiple new research

tasks must be engaged, including better estimates of fraud, new models and data on the externalities

of fraud, new and better instruments for estimating causal effects of anti-fraud regulation, and better

models and data on the chilling effects that financial regulation may have on legitimate activity.

Each task will be difficult, and each will likely require a separate stream of research—both

theoretical and empirical—before any plausible aggregate estimate of the benefits of any major

piece of financial regulation can be developed. While these tasks are being undertaken, regulation

will continue to be produced, modified, and criticized by political actors, and we conclude with

speculative suggestions on how policy-makers might improve on financial regulation from a

cost-benefit perspective without waiting on a fully realized research agenda.

Estimating the Incidence of Fraud and Its Direct Costs

The first task is to develop better methods of simply measuring the incidence of fraud and its

costs. This task is a necessary prerequisite before we can even roughly estimate the effect of

regulation aimed at reducing fraud. Many papers link fraud to equity compensation for executives (e.g., Burns and Kedia 2006; Johnson, Ryan, and Tian 2009) and non-independent or inexpert

corporate boards (e.g., Agrawal and Chadha 2005). Khanna, Kim, and Lu (2013) link fraud to CEO connectedness to other same-firm executives and directors. Dyck et al. (2010) find that multiple

10 On bank holding company window-dressing in the crisis generally, see Owens and Wu (2012).

SOX after Ten Years: A Multidisciplinary Review 655

Accounting Horizons September 2014

monitoring agents play a role in detecting major corporate frauds, including auditors, analysts, and the SEC, but also newspapers, employees, and non-financial regulators. Other papers examine whether

general measures of investor mood or beliefs are related to the overall incidence of fraud. 11

For the

most part, however, these studies only attempt to establish that relationships between incentives,

monitoring agents, or investor beliefs exist, and do not convincingly identify causality or present

evidence of how strong these relationships are, or what the overall incidence of corporate fraud is.

One difficulty confronting such studies is that, in observational or quasi-experiments set in the

real world, all concerned have incentives to hide fraud. 12

Fraud is already subject to criminal and

multiple civil penalties if discovered, so fraudsters have strong incentives to hide their activities. As

with crime generally, regulators and enforcement officials, too, have at least some incentives to

suggest that the incidence of undiscovered fraud is lower than it may actually be, since a large stock of

undiscovered fraud might be used to attack their efficacy. Victims have incentives both to over-report

fraud (in the wake of poor investment decisions, for example), but also to under-report it (because

there are social and psychological costs to acknowledging that one has been duped). Lev (2003)

reviews large sample studies of discontinuities in the distribution of public company earnings,

deviations from expected accruals, and patterns of earnings around suspicious circumstances, and

concludes that ‘‘this evidence suggests that the frequency of manipulations [of reported earnings] is

substantially higher than the 100–150 litigated or SEC-enforced cases per year.’’ As analyzed by

Feinstein (1990), partial observability presents challenges to empirical modeling.

Li (2013) and Wang (2013) use empirical models of both fraud incidence and detection that

allows for better inferences about the correlates of fraud (including both detected and undetected

fraud) to be drawn from detected frauds. 13

Each exploits the partial overlap in indicators of both

fraud incidence and fraud detection, using different econometric models. 14

Building on Wang’s (2013) work, Dyck, Morse, and Zingales (2013) exploit the failure of

Arthur Andersen that forced companies to change auditors, and on the assumption that new

auditors would ‘‘clean house’’ and reveal fraud that might otherwise have been unrevealed. They

estimate that the probability of a public company engaging in fraud in any given year is 15

percent—which would suggest a current stock of several hundred fraudulent public companies.

(They validate this measure with a survey of fraud observed by business school students at former

employers.) Dyck et al. (2013) estimate that the average corporate fraud generates losses of

between 22 percent and 40 percent of enterprise value (equity plus long-term debt) at a fraudulent

firm. In combination, these two estimates (0.15 3 0.22 ¼ 0.03) estimate an aggregate lower bound on the cost to investors in fraudulent firms of roughly 3 percent of enterprise value. Applied to the

market capitalization of all U.S. public companies, this implies a total as-yet-unrevealed direct cost

11 For example, Hertzberg (2013) and Povel, Singh, and Winton (2007) develop theories about whether and how the incidence of fraud should relate to investor optimism about business conditions. Hertzberg (2013) posits a positive relationship; Povel et al. (2007) argue instead for an inverted U shaped relationship, peaking when investors believe conditions are good, but not extremely good. Wang, Winton, and Yu (2010) examine a sample of IPOs to test which of these theories better fits the data, and find that the latter does.

12 In controlled experiments, by contrast, researchers can know with much greater certainty the actual incidence of fraud and the effects of different ‘‘regulatory’’ treatments (see, e.g., Guttentag, Porth, and Fraidin [2008], who study the effect of disclosure on fraud incidence in such an experiment). But the ability to generalize beyond such experimental research settings is unclear at best.

13 While the fraud prediction and detection models in these papers allow better assessment of fraud, concerns still remain about the exogeneity of the instruments used to model underlying fraud as distinct from detected fraud.

14 Wang (2013) finds R&D (but not capital expenditure), abnormally low accounting or stock returns, and high stock turnover make fraud more likely, while fraud is less likely if a firm faces more analyst coverage; fraud detection, by contrast, is increased by M&A and analyst coverage, but decreased by R&D expenditures. Li finds that fraud incidence is increasing in executive pay performance sensitivity and decreasing in institutional ownership, while fraud detection is increasing in qualified audit opinions and the SEC’s inflation-adjusted budget per public company.

656 Coates and Srinivasan

Accounting Horizons September 2014

to investors of over $500 billion. A reduction by just 1 percent in this incidence of fraud would

generate $5 billion of direct benefits to current investors—before accounting for the indirect costs

of fraud (discussed below).

Zakolyukina (2013) estimates undetected intentional earnings manipulation from a sample of

1,500 firms in the post-SOX period. She finds that the probability of detection is only 9 percent and

generates a loss of 11 percent to the firm’s CEO wealth if detected. The inference she draws is that 66

percent of her sample have rational incentives to manipulate earnings, and that the value-weighted

bias in stock prices across the sample firms is 16 percent—in line with the Dyck et al. (2013) estimate

of the direct investor losses caused by fraud. Finally, Karpoff, Lee, and Martin (2008a) provide

estimates of the costs borne by firms that engage in fraud, and their managers. Firms that

misrepresented in the pre-SOX period and were subject to SEC AAERs incur penalties of $24

million per firm, but lose 7.5 times more (on a present value basis) in the form of lower sales and

higher contracting and financing costs. For each dollar a firm inflated its market value, it lost

$4.08 once the misconduct was revealed—$0.36 due to SEC penalties and $3.83 to reputation.

Survey evidence in Dichev, Graham, Harvey, and Rajgopal (2013) suggests that around 20

percent of firms exploit GAAP to misrepresent reported performance in financial statements.

These studies, while helpful in at least bounding the incidence of certain types of fraud, are

limited in various ways and include different factors. Future research could aim at more

comprehensive measures of fraud, to include not only auditor-discoverable fraud (studied in Dyck

et al. [2013]) and earnings manipulation (studied in Lev [2003] and Zakolyukina [2013]), but also

frauds that fall outside the scope of audits, such as insider trading, self-dealing (as at Enron), fraud-

based compensation (as at Tyco), frauds involving third parties (as at WorldCom) or technically

GAAP-compliant but deceptive accounting choices (as arguably was true at Lehman). As an

example, Anginer et al. (2012) identify suspicious options backdating activity by identifying likely

opportunistic option grants that are made at unusually favorable exercise prices. This allows them to

identify a sample—an upper bound—of all possible backdating cases from where they identify

firms that were prosecuted by the SEC. At a minimum, we suggest a better comprehensive data-

collection effort by all enforcement officials, across all relevant regulatory agencies (SEC, CFTC,

CFPB, FTC, Fed, OCC, FDIC, DOJ, FINRA, PCAOB, state agencies, and AGs), to identify

aggregate trends in the incidence of fraud. A companion effort might be for the same agencies to

create a single point-of-entry web-based system for whistle-blowers, victims, and others to report

suspicious or potentially fraudulent activity, which could provide directional, if noisy, data on as-

yet unproven fraud.

Modeling and Measuring the Externalities of Fraud

If more work needs to be done on models of fraud incidence, the equally important task of

estimating the costs of fraud remains farther behind. Few researchers have systematically

attempted to study and measure the social costs of financial frauds. Without such estimates, an

assessment of a regulation that reduces fraud would remain qualitative. Research particularly

needs to be conducted on the extent to which fraud results not only in direct monetary net losses

to the parties involved (the focus of Dyck et al. [2013], discussed above), but in both

psychological costs and externalities.15 Psychological effects (e.g., fear, distrust, and stress) can

15 Fraud is criminalized in part for these reasons—direct remediable civil damages are not thought to be large enough to provide sufficient incentive for private actors to enforce optimally (Shavell 1986). But criminal sanctions are reserved for a small subset of frauds—those in which clear evidence is available ex post for frauds caused by individuals with specific intent, and the nature of fraud is such that such evidence is often unavailable. Section 24 of the Securities Act of 1933 imposes criminal liability for ‘‘willful’’ violations; see also section 32(a) of the Securities Exchange Act of 1934.

SOX after Ten Years: A Multidisciplinary Review 657

Accounting Horizons September 2014

result in tangible consequences, including drug addiction, job loss, reduced income, health

effects, and even suicide. In the context of securities fraud, Freshman (2012) finds elevated

levels of post-traumatic stress disorder and related behavioral effects among Madoff victims. On

the externalities of fraud, Lev (2003)’s sketch is a useful starting point for the following

categories: 16

(1) Fraud increases the cost of capital for all firms, due to reduced quality of financial

information and heightened expected fraud-related losses and reduced confidence in public

securities markets and in markets more generally. 17

(2) Fraud wastes investment caused by misallocation of resources caused by fraudulent signals

of the value of firms or whole industries, as in the telecom and Internet bubbles. 18,19

(3) Fraud destroys value through the (costly) acquisition by fraudulent companies of other

companies, followed by mismanagement or outright theft and bankruptcy, causing further

transaction costs of reorganizing still-viable companies once the fraud is uncovered. 20

(4) Fraud induces increased bonding, monitoring and precautionary costs by investors to avoid

fraud, such as for audit firms, independent directors, appraisers, analysts, regulatory and

enforcement agencies, and prisons. 21

(5) Fraud imposes costs imposed on third parties that are dependent on the victims of the

initial fraud (e.g., family, business partners, creditors, and communities). 22

For example, consider the likely size of such costs flowing from the Madoff scandal, which

imposed significant direct losses on over 15,000 individual investors, each of whom presumably

had an average of two dependents or heirs, and many of whom were co-investors and borrowers

16 Anderson (1999) presents a similar list of indirect effects of crime generally. He estimates the indirect costs—what he categorizes as ‘‘crime-induced production,’’ opportunity costs, and risks to life and health—are roughly double the value of victim-to-criminal property transfers, and when he counts the costs incurred by criminals, the total costs of crime is more than double the value of those transfers (see Anderson 1999, Table 7). In other words, the external effects of crime generally greatly exceed their direct effects. See also Velikonja (2013).

17 Jain, Kim, and Rezaee (2008) show that market liquidity deteriorated following Enron and related scandals, and improved after SOX’s adoption, and Giannetti and Wang (2013) show that revelation of corporate frauds in a state caused equity holdings of households in that state to fall, increasing the cost of capital for non-fraudulent firms. For a more general study of the effect of trust on finance, see Guiso, Sapienza, and Zingales (2008); see also Bonaccorsi di Patti (2009) (crime, including fraud, increases borrowing costs and increases capital constraints).

18 For a review of studies showing that corporate finance decisions driven by capital market prices, including prices that deviate from fundamental values (i.e., mispricing), see Baker (2009); see also Baker, Stein, and Wurgler (2003) (modeling and presenting evidence that bubbles affect corporate investment).

19 Kedia and Phillippon (2009) model investment decisions of firms during periods of fraud and find empirical support for their prediction that fraud and earnings management distort hiring and investment decisions of firms, leading to over-investment and excessive hiring during periods of suspicious accounting. This leads to misallocation of resources in the economy.

20 For a model of merger and acquisition activity driven by mispricing, see Shleifer and Vishny (2003); for estimates of the costs of bankruptcy, see, e.g., Bris, Welch, and Zhu (2006 ) (estimating range from 2 to 20 percent of firm assets resulting from formal bankruptcy).

21 As noted above, audit fees were rising prior to SOX, due to market-driven demand for increased scrutiny of financial statements following the scandals that led to SOX. See Asthana et al. (2009). Likewise, separate from SOX, the NYSE and the NASDAQ adopted tighter corporate governance requirements in response to Enron et al., which tightened the criteria for and likely increased the costs of recruiting independent directors.

22 For studies showing spillover effects of restatements, see Gleason, Jenkins, and Johnson (2008); Durnev and Mangen (2009) (spillovers on same-industry companies); Files and Gurun (2011) (spillovers on competitors, customers, or suppliers of the restating firm). Karpoff, Lee, and Martin (2008b) find that of 2,206 individuals identified as responsible for a set of 788 SEC and DOJ enforcement actions for financial misrepresentations, 93 percent lost their jobs, most suffered significant financial losses, and 28 percent faced criminal charges and penalties, including prison sentences averaging 4.3 years. Their results imply that the 788 frauds generated led to 2,656 prison-years, roughly totaling $79 million for prison costs alone, based on data from Sabol, West, and Cooper (2009) and Kyckelhahn (2010).

658 Coates and Srinivasan

Accounting Horizons September 2014

with yet others, or makers of charitable donations to non-profits. 23

To date, the liquidation of the

Madoff entities has generated over $700 million in expenses—all a pure loss to investors, over and

above the amounts stolen by Madoff himself. 24

As a broader example, consider how fraudulent

home loans (whether due to borrower fraud, lender fraud, or both) had ripple effects in the last

financial bubble, partly generated through leverage and intermediation, so the one fraudulent loan

would affect not only the immediate parties to the loan but also securitization lenders, sponsors, and

other related parties; collateralized debt obligation investors, sponsors, and related parties;

structured investment vehicle (SIV) investors, sponsors, and related parties; investors in the banks

that sponsored the SIVs; borrower-customers of those banks, whose capital constraints and

heightened risk-aversion following the crisis caused a withdrawal or increase in the cost of credit;

employees and customers of businesses that failed as a result of the capital constraints generated by

the banks’ losses; family members of those employees, and so on.

How do we translate these anecdotal examples into a more general method for estimating not

just the losses to fraud victims, but the more general knock-on effects of fraud on society as a

whole? In the context of SOX specifically, Graham, Litan, and Sukhtankar (2002) is an isolated

effort to estimate fraud’s social costs from stock market price changes, arguably induced by fraud.

They relate large-scale instances of fraud revelation (as in the period leading up to SOX) and the

direct costs they impose on investors on the fraudulent companies to models of the role of the

stock market in the macroeconomy. The authors draw on the discussion of ‘‘shocks’’ to the equity premium in the broader analysis in Reifschneider, Tetlow, and Williams (1999) of the ‘‘U.S./Fed

model’’ used by the Treasury Department and the Federal Reserve to model the U.S. economy for purposes of setting monetary policy. In another attempt to assess externalities from fraud,

Giannetti and Wang (2013) use brokerage data of a sample of retail investors across the U.S. and

show that upon the revelation of fraud in a company in a particular state, all households in the

state, not just the ones owning stocks of fraud firms, reduce their equity holdings. Households

also decrease their ownership in all firms, not just the ones where fraud is revealed. Firms

headquartered in the same state as a firm involved in fraud experience a decline in the number of

shareholders, especially retail ones. Some state firms also experience an increase in cost of capital

and lower valuations.

Graham et al. (2002) first develop a rough range of estimates of the direct costs to investors of

the pre-SOX scandals, ranging from $10 to $23 billion, based on assuming that between 25 percent

and 100 percent of the market decline from March 2002 to July 2002 was caused by those scandals.

They then combine these estimates with the U.S./Fed model’s prediction that investment would fall

0.8 percent per year in response to a 20 percent decline in stock market wealth, to estimate the first-

year impacts of the scandals on the economy as a whole, producing social estimates of the costs of

the scandal ranging from $19 billion to $57 billion. Graham et al.’s (2002) estimates would

understate actual costs if the impact on U.S. GDP lasted for more than one year.

For similar models of the social costs of financial crises, such as the recent one, see Haldane

(2010) (estimating that the recent crisis caused losses of between 90 percent and 350 percent of

world GDP); De-Ramon, Iscenko, Osborne, Straughan, and Andrews (2012) (estimating the recent

crisis caused losses of between 18 percent and 48 percent of U.K. GDP); and Posner and Weyl

(2013) (estimating the losses at between 1 percent and 20 percent of U.S. GDP). As can be seen

from the range of estimates from these studies, the models remain highly sensitive to assumptions,

and produce highly varying results. Even if one thinks that ‘‘ordinary’’ corporate fraud produces less dramatic effects on the financial system as a whole than was the case in the recent crisis,

23 See Picard (2013, Exhibit A, p. 5). For charities harmed by the Madoff scandal, see Weiss and Birkner (2008).

24 See Picard (2013, Exhibit A, p. 2).

SOX after Ten Years: A Multidisciplinary Review 659

Accounting Horizons September 2014

considerably more work remains to be done to be able to estimate the social costs of financial fraud

in this way.

Finally, research on corporate fraud could learn from research on crime generally, which uses

several families of methods (see Ludwig [2010] and Donohue [2009] for overviews): (1) ‘‘hedonic’’

models in which changes or variation in market prices affected by crime to infer social costs (e.g.,

Thaler 1978; Hoehn, Berg, and Blomquist 1987; Viscusi 2000); (2) the use of surveys of

willingness-to-pay by potential victims for a reduction in the odds of crime (e.g., Cohen, Rust,

Steen, and Tidd 2004; Nagin, Piquero, Scott, and Steinberg 2006 ); (3) estimating each of the direct

and indirect effects separately and aggregating them (e.g., Anderson 1999); and (4) relating

responses to surveys of crime victims to respondent wealth or income and inferring a ‘‘shadow

price’’ for the effects of crime (e.g., Moore and Shepherd 2006 ). 25

Each method has limitations and

weaknesses, and is probably best used in combination to produce cost ranges, as in Donohue

(2009).

Estimating Causal Effects of Financial Regulation

With a better framework for estimating the incidence and costs of fraud in hand, researchers

could then better estimate the benefits of specific legal or regulatory changes, such as reflected in

SOX. To date, however, most studies of SOX (as reflected in our discussion above), have not used

research designs well adapted for this purpose, and instead use simple before-and-after comparisons

that fail to control for contemporaneous changes in the objects of study. Studying such changes is

much harder than rocket science, which after all models relatively simple inert objects moving

through space, and not large groups of independent agents interacting in a complex regulatory and

market environment. Given the complexity of the task and the inevitably low power of such studies,

we continue to need to find better quasi-experimental designs to better control adequately for

omitted variables and identify causal effects.

Better are difference-in-difference designs, such as those used by Kamar et al. (2009)

(studying going private), Johnson and Madura (2009) (studying going public); Piotroski and

Srinivasan (2008) (studying foreign listings); Kang et al. (2010) and Bargeron et al. (2010)

(studying investment decisions). Even those studies can also be misleading if factors affect events

in the nominal control sample in a similar fashion as in the nominally ‘‘treated’’ sample, as seems

to be true for the investment decision studies. Better still are discontinuity designs, used by Iliev

(2010) (studying changes in audit costs and accounting quality); Albuquerque and Zhu (2013)

(studying investment decisions), but the findings of such studies may not generalize beyond the

immediate discontinuity in question. Perhaps best of all are time-series designs studying multiple

events, such as those used by Leuz et al. (2008) (several events in the phase-in and extended

exemptions for small firms from compliance with section 404) and Arping and Sautner (2013)

(staged phase-in for foreign firms cross-listed in the U.S. to comply with section 404). But it is

rare to find multiple, objective, and independent events in legal or accounting rule-systems with

related implications.

25 An open conceptual issue is whether to count the utility of a criminal in estimating social welfare effects of crime. Assume, for example, a fraudster obtains $1 from a victim and spends it on food. Is the direct social loss $0 or $1? If the criminal’s utility is ignored and the fraud has no effect besides the transfer of $1, the social loss is $1. If the criminal’s utility is counted equally with the victim’s, and neither attaches unusual utility to the $1, the social loss is $0. As outlined above, such fraud generates other losses, such as precautionary expenditures and externalities, but the question of how to count gains or benefits to fraudsters remains disputed. Compare Cook (1983) (criminals’ utility should count) with Cohen (2005) and Ludwig (2006 ) (it should not count). Anderson (1999)’s data suggest the difference is significant.

660 Coates and Srinivasan

Accounting Horizons September 2014

Modeling and Measuring Chilling Effects of Financial Regulation

Just as better causal designs are needed to study the effects of regulatory changes, so too are

better designs for studying their costs. This is true even of direct costs, given the incentives of

affected companies to exaggerate those costs in public comments, and the fact that most regulatory

burdens will have a fixed, up-front component and a smaller ongoing cost, and can be reduced over

time as compliance technologies improve. But the need is even greater with respect to indirect costs

such as changes in risk-taking and investment, which can plausibly dwarf the direct costs in

magnitude. SOX, for example, is said to have caused changes in the risk of personal liability facing

managers and directors and in the risk of reputational harms and opportunity costs created by

litigation. If true, difficult-to-explain and legitimate business risks may be foregone, firms may

decline to go public or otherwise avoid the burdens of the law, with resulting social costs. These

costs are likely to affect more firms, including firms not yet created, than the out-of-pocket costs of

compliance at the time a law is initially adopted. Better specification and measurement of indirect

costs, and better causal designs to estimate their magnitudes, will reduce the uncertainty associated

with the net effects of any given regulatory change.

Regulatory Experiments

The difficulty in assessing causal effects of regulation arises from our inability to conduct a

randomized trial. As discussed earlier, the staggered implementation for companies of different

sizes and size cut-off for Section 404 applicability have provided the best experimental settings

for causal analysis of costs and benefits of SOX. Prior experience suggests a few useful

approaches to improve our ability to assess costs and benefits with some degree of causal

inference. The SEC, in implementing Regulation SHO (regulation of short selling effective

January 2005), adopted a randomized testing approach in allowing a temporary suspension of

some short-sale restrictions ( price tests) for a set of roughly one thousand U.S. stocks—the

‘‘Pilot’’ stocks were randomly selected to be every third from the Russell 3000 index ranked by

trading volume. Diether, Lee, and Werner (2009) use this randomization to assess the

effectiveness of price tests and find that the rules introduce distortions in order flow, thus

affecting market quality. Similar randomization in implementation of important rules for large

enough samples can allow for causal inferences.

A second approach is to allow voluntary opt-in or opt-out of regulations. While the voluntary

nature of the opting decision introduces endogeneity concerns, this method allows for less costly

trials since companies that find regulations costly are more likely to opt out. For example,

Szewczyk and Tsetsekos (1992) examine Pennsylvania Senate Bill 1310 that introduced sweeping

anti-takeover protection but allowed companies the option to reject or retain provisions of state

level anti-takeover provisions. They find an abnormal stock return of about negative 9 percent from

introduction to enactment of the bill and a positive abnormal two-day return of 0.65 percent for

firms that rejected all provisions of the bill. Manzi (2012) discusses several experiments in Welfare

Reform in the U.S. where individual states were allowed to opt out of federal regulations in return

for data collection on several outcomes, which eventually led to scientific evidence on the efficacy

of the programs.

What to Do in the Meantime?

Here we return to the fact that many costs (direct costs) are easier to observe and quantify (even

if such estimates are likely to be overstated), while other costs (opportunity, risk-taking) are hard to

observe and quantify, and the benefits being hardest to estimate of all. This raises a methodological

challenge for policy-makers: if they are truly neutral on the net costs/benefits coming to the

SOX after Ten Years: A Multidisciplinary Review 661

Accounting Horizons September 2014

analysis, they have to implicitly weight the results of quantitative assessments differently, given the

difference in ease of quantifying costs/benefits.

One way to address this imbalance might be to develop a ‘‘multiplier’’ for fraud’s costs,

and for the benefits of reducing fraud through any given regulatory change. This multiplier

would be similar to the multiplier used to estimate the effects of government spending on

economic growth—with all of the controversy that the analogy suggests, whatever the result of

the debate over the size of the multiplier, it would be better than either accepting zero as the

implicit number for difficult-to-quantify benefits of anti-fraud regulation within a cost-benefit

framework, or to decline to pursue quantification altogether, relying solely on faulty and

politically influenced regulatory judgments of elected lawmakers and appointed regulatory

officials.

Alternatively, policy-makers can put together and refine but rely on admittedly crude

assumptions such as those used by Dyck et al. (2013) and Graham et al. (2002), discussed above.

Such methods will create ballpark estimates of the benefits of tightened financial regulations, which

can then be used to compare to equally crude estimates of such regulations’ costs. Even crude

efforts at quantification of the full range of regulatory costs and benefits will at least focus public

and policy debate on common questions. Without it, the result is likely to be continued polarization

over financial regulations’ effects, and continued vacillation between excessively tight and

excessively permissive laws.

CONCLUSION

Coming in the aftermath of several large-scale frauds in the early 2000s, SOX sought to

empower regulators, auditors, and corporate boards to improve governance and reduce fraud. With

the passage now of over ten years, there is substantial research on the consequences of the Act. We

assess the impact of SOX using findings from scholarly research to help researchers understand the

strengths and limitations of current research in informing about significant regulatory interventions,

as well as to inform regulators and policy-makers on the design of future regulations or changes to

current ones.

We find that despite its controversial nature, the Act has survived almost intact since its

enactment. The PCAOB has become an important part of the regulatory landscape, having

survived a high profile lawsuit that challenged its constitutionality. Our survey suggests that some

of the early concerns about SOX were likely overblown. There is scant evidence of

‘‘federalization’’ of corporate law. Significant modifications to the implementation of SOX’s

core provision relating to internal controls (section 404) have made it less costly to implement

and hence more acceptable.

Several studies that have estimated the direct costs of SOX (especially section 404)

implementation find that these costs fall disproportionately on smaller firms. Modifications to

section 404 and subsequently the JOBS Act were aimed at lowering the cost imposed on smaller

firms. In terms of indirect costs, we conclude that the evidence of costs is not clear cut. While

several smaller firms left the public equity market and smaller foreign firms shunned the U.S. as a

listing venue, these firms fit the profile of firms more susceptible to fraud (and also reflect reversal

from a higher propensity in the mid-1990s for smaller firms to access public equity markets).

Hence, the social welfare implication of implicitly deterring these smaller firms from public markets

is unclear. A more serious concern is the documented decline in investments-to-capital ratio in the

U.S. post-SOX, with implications for innovation and risk taking in the U.S. economy. However,

research also documents that the trend toward lower investments started prior to SOX and the

post-SOX trend is merely a continuation of a prior pattern. We conclude that, while there is some

662 Coates and Srinivasan

Accounting Horizons September 2014

evidence on indirect costs, research presently has not produced robust evidence on the cause or the

magnitude of the effect.

Research has documented benefits in financial reporting quality after SOX. While there is a

significant time period effect, there is less evidence on the causality. The staggered implementation

of SOX 404 allowed researchers to produce some causal evidence that section 404 implementation

has produced financial reporting benefits. There is also significant evidence that internal control

weakness reports produced as a result of section 404 are valuation-relevant for investors and prompt

companies to make managerial and governance improvements. Overall, there appears to have been

an improvement in financial reporting quality after SOX, some of which can be attributed to section

404.

We conjecture that the policy opposition to SOX is driven to a large extent by the observable

direct costs. Estimating the benefits and indirect costs are difficult tasks, made even harder by the

difficulty of drawing causal inferences about the Act’s effects. Further, without a better handle on

the externalities created by frauds, it is hard to assess costs and benefits of regulation designed to

combat fraud. We make several suggestions in the short and long term that researchers and

policy-makers can adopt. The first task is to estimate the extent of the incidence of fraud (not just

fraud that is detected) and then to model and measure the externalities caused by fraud. The next

task would be to estimate the causal effects of regulation aimed at combating fraud. And the third

task is to model and measure the negative effects of regulation. Together these tasks are challenging

but critical to ensure policy making is well informed by research.

REFERENCES

Abbott, L. J., K. A. Gunny, and T. C. Zhang. 2013. When the PCAOB talks, who listens? Evidence from

stakeholder reaction to GAAP-deficient PCAOB inspection reports of small auditors. Auditing: A

Journal of Practice & Theory 32 (2): 1–31.

Abbott, L. J., S. Parker, and G. F. Peters. 2004. Audit committee characteristics and restatements. Auditing: A Journal of Practice & Theory 23 (1): 69–87.

Agrawal, A., and S. Chadha. 2005. Corporate governance and accounting scandals. Journal of Law and Economics 48: 371–406.

Akhigbe, A., and A. Martin. 2006. Valuation impact of Sarbanes-Oxley: Evidence from disclosure and

governance within the financial services industry. Journal of Banking and Finance 30: 989–1006.

Albuquerque, A. M., and J. L. Zhu. 2013. Has Section 404 of the Sarbanes-Oxley Act Discouraged Corporate Risk-Taking? New Evidence from a Natural Experiment. Working paper, Boston

University. Available at: http://ssrn.com/abstract¼1997228 Alexander, C. R., S. W. Bauguess, G. Bernile, Y. A. Lee, and J. Marietta-Westberg. 2013. Economic effects

of SOX Section 404 compliance: A corporate insider perspective. Journal of Accounting and

Economics 56 (2-3): 267–290.

Altamuro, J. L., and A. L. Beatty. 2010. How does internal control regulation affect financial reporting?

Journal of Accounting and Economics 49 (1–2): 58–74.

Anderson, D. 1999. The aggregate burden of crime. Journal of Law and Economics 42 (2) (October): 611–

642.

Anginer, D., M. P. Narayanan, C. Schipani, and N. Seyhun. 2012. Should size matter when regulating

firms? Implications from backdating of executive options. N.Y.U. Journal of Legislation and Public Policy 15: 1–44.

Arping, S., and Z. Sautner. 2013. Did SOX section 404 make firms less opaque? Evidence from cross-listed

firms. Contemporary Accounting Research 30: 1133–1165.

Ashbaugh-Skaife, H., R. LaFond, and B. W. Mayhew. 2003. Do nonaudit services compromise auditor

independence? Further evidence. The Accounting Review 78 (3): 611–639.

SOX after Ten Years: A Multidisciplinary Review 663

Accounting Horizons September 2014

Ashbaugh-Skaife, H., D. W. Collins, W. R. Kinney, and R. LaFond. 2008. The effect of SOX internal

control deficiencies and their remediation on accrual quality. The Accounting Review 83: 217–250.

Ashbaugh-Skaife, H., D. W. Collins, W. R. Kinney, and R. LaFond. 2009. The effect of SOX internal

control deficiencies on firm risk and cost of equity. Journal of Accounting Research 47 (1): 1–43.

Asthana, S., S. Balsam, and S. Kim. 2009. The effect of Enron, Andersen, and Sarbanes-Oxley on the

market for audit services. Accounting Research Journal 22 (1).

Baker, M., J. Stein, and J. Wurgler. 2003. When does the market matter? Stock prices and the investment of

equity-dependent firms. Quarterly Journal of Economics: 969–1005.

Baker, M. 2009. Capital market-driven corporate finance. Annual Review of Financial Economics 1 (1): 181–205.

Bargeron, L., K. Lehn, and C. Zutter. 2010. Sarbanes-Oxley and corporate risk-taking. Journal of Accounting and Economics 49: 34–52.

Bartlett, R. 2006. Going private but staying public: Reexamining the effect of Sarbanes-Oxley on firms’

going-private decisions. Chicago Law Review 76: 7–43.

Bartov, E., and D. Cohen. 2009. The ‘‘numbers game’’ in the pre- and post-Sarbanes-Oxley eras. Journal of Accounting, Auditing, and Finance 24 (4): 505–534.

Bates, T., K. Kahle, and R. Stulz. 2009. Why do U.S. firms hold so much more cash than they used to?

Journal of Finance 64: 1985–2022.

Bazerman, M. H., and D. A. Moore. 2011. Is it time for auditor independence yet? Accounting, Organizations and Society (December).

Bebchuk, L. A., J. Bachelder, R. C. Campos, B. S. Georgiou, A. G. Hevesi, W. Lerach, R. Mendelsohn, R.

Monks, T. Myerson, J. F. Olson, L. Strine, and J. C. Wilcox. 2006. Director liability. Delaware Journal of Corporate Law 31 (3): 1011–1045.

Bhattacharya, U., P. Groznik, and B. Haslem. 2007. Is CEO certification of earnings numbers value-

relevant? Journal of Empirical Finance 14: 611–635.

Black, B., B. Cheffins, and M. Klausner. 2006. Outside director liability. Stanford Law Review 58: 1055– 1160.

Bonaccorsi di Patti, E. 2009. Weak Institutions and Credit Availability: The Impact of Crime on Bank Loans. Working paper, Bank of Italy. (July 10) Available at: http://ssrn.com/abstract¼1606242

Bova, F., M. Minutti-Meza, G. Richardson, and D. Vyas. 2014. The Sarbanes-Oxley Act and exit strategies

of private firms. Contemporary Accounting Research (forthcoming).

Brau, J. C., and S. E. Fawcett. 2006. Initial public offerings: An analysis of theory and practice. The Journal of Finance 61: 399–436.

Bris, A., I. Welch, and N. Zhu. 2006. The costs of bankruptcy: Chapter 7 liquidation versus Chapter 11

reorganization. Journal of Finance 61 (June): 1253–1303.

Brochet, F. 2010. Information content of insider trades before and after the Sarbanes-Oxley Act. The Accounting Review 85 (2): 419–446.

Brochet, F., and S. Srinivasan. 2013. Accountability of independent directors—Evidence from firms subject

to securities litigation. Journal of Financial Economics 111 (2): 430–449.

Burch, R. F. 2006. Director oversight and monitoring: The standard of care and the standard of liability

post-Enron. Wyoming Law Review 6: 482.

Burks, J. J. 2010. Disciplinary measures in response to restatements after Sarbanes-Oxley. Journal of

Accounting and Public Policy 29 (3): 195–225.

Burks, J. J. 2011. Are investors confused by restatements after Sarbanes-Oxley? The Accounting Review 86 (2): 507–539.

Burns, N., and S. Kedia. 2006. The impact of performance-based compensation on misreporting. Journal of Financial Economics 79: 35–67.

Butler, H. N., and L. E. Ribstein. 2006. The Sarbanes-Oxley Debacle: How to Fix it and What We’ve

Learned. Washington, D.C.: American Enterprise Institute Press.

Carter, M. E., L. J. Lynch, and S. C. Zechman. 2009. Changes in bonus contracts in the post-Sarbanes-

Oxley era. Review of Accounting Studies 14 (4).

664 Coates and Srinivasan

Accounting Horizons September 2014

Center for Audit Quality (CAQ). 2010., The CAQ’s Fourth Annual Individual Investor Survey. (September). Washington, DC: CAQ.

Charles River Associates (CRA). 2005. Sarbanes-Oxley Section 404: Costs and Remediation of Deficiencies. Available at: https://www.sec.gov/spotlight/soxcomp/soxcomp-all-attach.pdf

Cheng, B., S. Srinivasan, and G. Yu. 2013. Securities litigation risk for foreign companies listed in the U.S.

Available at: http://ssrn.com/abstract¼2163864 Cheng, M., D. S. Dhaliwal, and Y. Zhang. 2013. Does investment efficiency improve after the disclosure of

material weaknesses in internal control over financial reporting? Journal of Accounting and Economics 56: 1–18.

Chhaochharia, V., and Y. Grinstein. 2007. Corporate governance and firm value: The impact of the 2002

governance rules. Journal of Finance 62: 1789–1825.

Chung, H., and S. Kallapur. 2003. Client importance, nonaudit services, and abnormal accruals. The Accounting Review 78 (4): 931–955.

Coates IV, J. C. 2007. The goals and promise of the Sarbanes-Oxley Act. Journal of Economic Perspectives 21: 91–116.

Coates IV, J. C. 2014. Cost-Benefit Analysis of Financial Regulation: Case Studies and Implications.

Working paper, Harvard University. Available at: http://ssrn.com/abstract¼2375396 Coffee, J. C., Jr. 1999. The future as history: The prospects for global convergence in corporate governance

and its implications. Northwestern University Law Review 93: 641–708.

Cohen, M., R. Rust, S. Steen, and S. Tidd. 2004. Willingness-to-pay for crime control programs.

Criminology 42: 86–106.

Cohen, M. 2005. The Costs of Crime and Justice. New York, NY: Routledge.

Cohen, D., A. Dey, and T. Lys. 2008. Real and accrual-based earnings management in the pre- and post-

Sarbanes-Oxley periods. The Accounting Review 83: 757–787.

Cohen, J. R., C. Hayes, G. Krishnamoorthy, G. S. Monroe, and A. Wright. 2013. The effectiveness of SOX

regulation: An interview study of corporate directors. Behavioral Research in Accounting 25 (1).

Committee on Government Reform. 2006. A Balancing Act: Cost, Compliance and Competitiveness after Sarbanes-Oxley. Hearing 109-217, June 19. Washington, DC: Government Printing Office.

Cook, P. J. 1983. Costs of crime. In Encyclopedia of Crime and Justice, edited by S. H. Kadish. New York, NY: Free Press.

Costello, A. M., and R. Wittenberg-Moerman. 2011. The impact of financial reporting quality on debt

contracting: Evidence from internal control weakness reports. Journal of Accounting Research 49 (1).

Cox, J. D., and R. S. Thomas. 2005. Public and private enforcement of the securities laws: Have things

changed since Enron? Notre Dame Law Review 80.

Cox, J. D. 2013. Strengthening financial reporting: An essay on expanding the auditor’s opinion letter.

George Washington Law Review 81: 1036–1061.

Daniel, N., D. Denis, and L. Naveen. 2008. Do firms manage earnings to meet dividend thresholds? Journal of Accounting and Economics 45 (March): 2–26.

Dechow, P. M., W. Ge, C. R. Larson, and R. G. Sloan. 2011. Predicting material accounting misstatements.

Contemporary Accounting Research 28: 17–82.

Dee, C. C., A. Lulseged, and T. Zhang. 2011. Client stock market reaction to PCAOB sanctions against a

Big 4 auditor. Contemporary Accounting Research 28 (1): 263–291.

DeFond, M. L., K. Raghunandan, and K. R. Subramanyam. 2002. Do non-audit service fees impair auditor

independence? Evidence from going concern audit opinions. Journal of Accounting Research 40 (4): 1247–1274.

DeFond, M. L., R. N. Hann, and X. Hu. 2005. Does the market value financial expertise on audit

committees of boards of directors? Journal of Accounting Research 43: 153–193.

DeFond, M. L., and C. S. Lennox. 2011. The effect of SOX on small auditor exits and audit quality. Journal of Accounting and Economics 52: 21–40.

DeFond, M. L., and J. Zhang. 2013. A review of archival auditing research. Journal of Accounting and Economics (forthcoming).

SOX after Ten Years: A Multidisciplinary Review 665

Accounting Horizons September 2014

De-Ramon, S., Z. Iscenko, M. Osborne, M. Straughan, and P. Andrews. 2012. Measuring the impact of

prudential policy on the macroeconomy: A practical application to Basel III and other responses to

the financial crisis. FSA Occasional Paper Series 42 (May). Desai, H., S. Rajgopal, and J. Yu. 2013. Did information intermediaries see the warning signals of the

banking crisis from leading indicators in banks’ financial statements? Contemporary Accounting Research (forthcoming).

Dhaliwal, D. S., V. Naiker, and F. Navissi. 2010. The association between accruals quality and the

characteristics of accounting experts and mix of expertise on audit committees. Contemporary Accounting Research (Fall).

Dichev, I., J. Graham, C. Harvey, and S. Rajgopal. 2013. Earnings quality: Evidence from the field. Journal of Accounting and Economics 56 (1–2/Supplement 1): 1–33.

Diether, M., J. Lee, and J. Werner. 2009. It’s SHO time! Short-sale price tests and market quality. The Journal of Finance 64: 37–73.

Doogar, R., P. Sivadasan, and I. Solomon. 2010. The regulation of public company auditing: Evidence from

the transition to AS5. Journal of Accounting Research 48 (4): 795–814. Doogar, R., S. Rowe, and P. M. Sivadasan. 2013. Asleep at the wheel (again)? Evidence from bank audits

during the financial crisis. Contemporary Accounting Research (forthcoming). Donohue, J. 2009. Assessing the relative benefits of incarceration: The overall change over the previous

decades and the benefits on the margin. In Do Prisons Make Us Safer? The Benefits and Costs of the Prison Boom, edited by S. Raphael, and M. Stoll. New York, NY: Russell Safe Foundation.

Doyle, J., W. Ge, and S. McVay. 2007. Accruals quality and internal control over financial reporting. The Accounting Review 82 (5): 1141–1170.

Durnev, A., and C. Mangen. 2009. Corporate investments: Learning from restatements. Journal of Accounting Research 47. Available at: http://www.artdurnev.com/wp-content/uploads/2010/06/ DurnevMangen.pdf

Dyck, A., A. Morse, and L. Zingales. 2010. Who blows the whistle on corporate fraud? Journal of Finance 65: 2213–2253.

Dyck, A., A. Morse, and L. Zingales. 2013. How pervasive is corporate fraud? Working paper, Rotman

School of Management.

Engel, E., R. Hayes, and X. Wang. 2007. The Sarbanes-Oxley Act and firms’ going-private decisions.

Journal of Accounting and Economics 44: 116–145. Feinstein, J. S. 1990. Detection controlled estimation. Journal of Law and Economics 33 (1): 233–276. Feng, M., C. Li, and S. McVay. 2009. Internal control and management guidance. Journal of Accounting

and Economics 48 (2–3): 190–209. Ferola, P. 2007. The role of audit committees in the wake of corporate federalism: Sarbanes-Oxley’s creep

into state corporate law. Journal of Business and Securities Law 7: 143–160. Files, R., and U. Gurun. 2011. Lenders’ Response to Restatements along the Supply Chain. Working paper,

The University of Texas at Dallas. Available at: http://ssrn.com/abstract¼1636862 Financial Executives Research Foundation (FERF). 2005. Special Survey on Sarbanes-Oxley Section 404

Implementation. Morristown, NJ: Financial Executives International and Financial Executives Research Foundation.

Freshman, A. 2012. Financial disaster as a risk factor for posttraumatic stress disorder: Internet survey of

trauma in victims of the Madoff Ponzi scheme. Health and Social Work. Gao, F., J. Wu, and J. Zimmerman. 2009. Unintended consequences of granting small firms exemptions

from securities regulation: Evidence from the Sarbanes-Oxley Act. Journal of Accounting Research 47: 459–506.

Gao, Y. 2011. The Sarbanes-Oxley Act and the choice of bond market by foreign firms. Journal of Accounting Research 49 (4): 933–968.

Gao, X., J. R. Ritter, and Z. Zhu. 2013. Where have all the IPOs gone? Available at: http://ssrn.com/

abstract¼1954788 Giannetti, M., and T. Y. Wang. 2013. Corporate scandals and household stock market participation.

Available at: http://ssrn.com/abstract¼2331588

666 Coates and Srinivasan

Accounting Horizons September 2014

Gleason, C. A., N. T. Jenkins, and W. B. Johnson. 2008. The contagion effects of accounting restatements.

The Accounting Review 83: 83–110.

Government Accountability Office (GAO). 2013. SEC Should Consider Requiring Companies to Disclose Whether They Obtained an Auditor Attestation. GAO-13-582. (July). U.S. Government Account- ability Office, Report to Congressional Committees, Internal Controls. Washington, DC: GAO.

Graham, C., R. Litan, and S. Sukhtankar. 2002. The Bigger They Are, the Harder They Fall: An Estimate of the costs of the Crisis in Corporate Governance. Working paper, Brookings Institution.

Greenspan, A. 2003. Testimony before the Committee on Financial Services. U.S. House of

Representatives. (July 15). Washington, DC: Government Printing Office.

Griffin, P. A., and D. H. Lont. 2005. Taking the oath: Investor response to SEC certification under

Sarbanes-Oxley. Journal of Contemporary Accounting and Economics 1: 27–63.

Griffin, P. A., and D. H. Lont. 2007. An analysis of the audit fees following the passage of Sarbanes-Oxley.

Asia Pacific Journal of Accounting and Economics 14 (2): 161–192.

Guiso, L., P. Sapienza, and L. Zingales. 2008. Trusting the stock market. Journal of Finance 63 (6 ): 2557– 2600.

Guttentag, M., C. Porth, and S. Fraidin. 2008. Brandeis’ policeman: Results from a laboratory experiment

on how to prevent corporate fraud. Journal of Empirical Legal Studies 5: 239–273.

Haldane, A. 2010. The $100 billion question. Speech by Bank of England Executive Director of Financial

Stability, March. Available at: http://www.bis.org/review/r100406d.pdf

Hammersley, J., L. Myers, and C. Shakespeare. 2008. Market reactions to the disclosure of internal control

weakness and to the characteristics of those weaknesses under section 302 of the Sarbanes Oxley Act

of 2002. Review of Accounting Studies 13 (1): 141–65.

Hennes, K. M., A. J. Leone, and B. P. Miller. 2008. The importance of distinguishing errors from

irregularities in restatement research: The case of restatements and CEO/CFO turnover. The Accounting Review 83 (6 ): 1487–1519.

Hertzberg, A. 2013. Managerial Incentives, Misreporting, and the Timing of Social Learning: A Theory of Slow Booms and Rapid Recessions. Working paper, Columbia Business School.

Hochberg, Y., P. Sapienza, and A. Vissing-Jorgensen. 2009. A lobbying approach to evaluating the

Sarbanes-Oxley Act of 2002. Journal of Accounting Research 47: 519–583.

Hoehn, J., M. Berger, and G. Blomquist. 1987. A hedonic model of interregional wages, rents, and amenity

values. Journal of Regional Science 27 (4): 605–620.

Hostak, P., E. Karaoglu, T. Lys, and G. Yang. 2013. An examination of the impact of the Sarbanes-Oxley

Act on the attractiveness of U.S. capital markets for foreign firms. Review of Accounting Studies 18

(2): 522–559.

Iliev, P. 2010. The effect of SOX section 404: Costs, earnings quality, and stock prices. Journal of Finance

45: 1163–1196.

IPO Task Force. 2011. Rebuilding the IPO On-Ramp: Putting Emerging Companies and the Job Market

Back to the Road on Growth. (October 20). Available at: http://www.sec.gov/info/smallbus/acsec/ rebuilding_the_ipo_on-ramp.pdf

Jain, P., and Z. Rezaee. 2006. The Sarbanes-Oxley Act of 2002 and capital-market behavior: Early

evidence. Contemporary Accounting Research 23: 629–654.

Jain, P., J. Kim, and Z. Rezaee. 2008. The Sarbanes-Oxley Act of 2002 and market liquidity. Financial Review 43 (3): 361–382.

Jennings, J. N., S. Kedia, and S. Rajgopal. 2013. The deterrent effects of SEC enforcement and class action

litigation on peer firms’ earnings quality. Working paper, Goizueta Business School and Emory

University.

Jiang, J., K. Petroni, and I. Wang. 2010. CEOs and CFOs: Who have more influence on earnings

management? Journal of Financial Economics 96 (3): 513–526.

Jin, J., K. Kanagaretnam, and G. Lobo. 2011. Ability of accounting and audit quality variables to predict

banking failure during the financial crisis. Journal of Banking and Finance 35: 2811–2819.

Johnson, J., and J. Madura. 2009. The pricing of IPOs post-Sarbanes-Oxley. Financial Review 44: 291–310.

SOX after Ten Years: A Multidisciplinary Review 667

Accounting Horizons September 2014

Johnson, S., H. Ryan, and Y. Tian. 2009. Managerial incentives and corporate fraud: The sources of

incentives matter. Review of Finance 13: 115–145.

Johnstone, K., C. Li, and K. Rupley. 2011. Changes in corporate governance associated with the revelation

of internal control material weaknesses and their subsequent remediation. Contemporary Accounting Research 28 (1): 331–383.

Jones, R. 2004. Rethinking corporate federalism in the era of corporate reform. Journal of Corporate Law 29: 625–663.

Kahan, M., and E. B. Rock. 2011. The insignificance of proxy access. Virginia Law Review 97: 1347.

Kamar, E., P. Karaca-Mandic, and E. Talley. 2009. Going-private decisions and the Sarbanes–Oxley Act of

2002: A cross-country analysis. Journal of Law, Economics and Organization 25: 107–133.

Kang, Q., Q. Liu, and R. Qi. 2010. The Sarbanes-Oxley Act and corporate investment: A structural

assessment. Journal of Financial Economics 96: 291–305.

Karpoff, J., D. Lee, and G. Martin. 2008a. The costs to firms of cooking the books. Journal of Financial and Quantitative Analysis 43: 581–612.

Karpoff, J., D. Lee, and G. Martin. 2008b. The consequences to managers for financial misrepresentation.

Journal of Financial Economics 88: 193–215.

Kedia, S., and T. Phillippon. 2009. The economics of fraudulent accounting. Review of Financial Studies 22: 2169–2198.

Kedia, S., and S. Rajgopal. 2011. Do the SEC’s enforcement preferences affect corporate misconduct?

Journal of Accounting and Economics 51 (3): 257–278.

Kedia, S., K. Koh, and S. Rajgopal. 2013. Evidence on Contagion in Corporate Misconduct. Working paper, Emory University.

Khanna, V., E. H. Kim, and Y. Lu. 2013. CEO Connectedness and Corporate Frauds. Working paper, University of Michigan and Tsinghua University. Available at: http://ssrn.com/abstract¼2323251

Kim, J.-B., B. Y. Song, and L. Zhang. 2011. Internal control weakness and bank loan contracting: Evidence

from SOX section 404 disclosures. The Accounting Review 86 (4): 1157–1188.

Kim, E. H., and Y. Lu. 2013. Corporate governance reforms around the world and cross-border acquisitions.

Journal of Corporate Finance 22: 236–253.

Kinney, W. R., Z.-V. Palmrose, and S. Scholz. 2004. Auditor independence, Non-audit services, and

restatements: Was the U.S. government right? Journal of Accounting Research 42 (3): 561–588.

Klein, A. 2002. Audit committee, board of director characteristics, and earnings management. Journal of Accounting and Economics 33: 375–400.

Koh, K., D. Matsumoto, and S. Rajgopal. 2008. Meeting or beating analyst expectations in the post-

scandals world: Changes in stock market rewards and managerial actions. Contemporary Accounting Research 25: 1067–1098.

Koh, K., S. Rajgopal, and S. Srinivasan. 2013. Non-audit services and financial reporting quality: Evidence

from 1978–1980. Review of Accounting Studies 18: 1–33.

Kyckelhahn, T. 2010. Justice Expenditures and Employment, FY 1982–2007–Statistical Tables. Bureau of Statistics. Available at: http://www.bjs.gov/content/pub/pdf/jee8207st.pdf

Langevoort, D. 2007. The social construction of Sarbanes-Oxley. Michigan Law Review 105: 1817–1855.

Lennox, C., and J. Pittman. 2010. Auditing the auditors: Evidence on the recent reforms to the external

monitoring of audit firms. Journal of Accounting and Economics 49: 84–103.

Leuz, C. 2007. Was the Sarbanes-Oxley Act of 2002 really this costly? A discussion of evidence from event

returns and going-private decisions. Journal of Accounting and Economics 44: 146–165.

Leuz, C., A. Triantis, and T. Wang. 2008. Why do firms go dark? Causes and economic consequences of

voluntary SEC deregistrations. Journal of Accounting and Economics 45: 181–208.

Lev, B. 2003. Corporate earnings: Facts and fiction. Journal of Economic Perspectives 17: 27–50.

Li, H., M. Pincus, and S. Rego. 2008. Market reaction to events surrounding the Sarbanes-Oxley Act of

2002. Journal of Law and Economics 51: 111–134.

Li, C., L. Sun, and M. Ettredge. 2010. Financial executive qualifications, financial executive turnover, and

adverse SOX 404 Opinions. Journal of Accounting and Economics 50: 93–110.

668 Coates and Srinivasan

Accounting Horizons September 2014

Li, S. 2013. Corporate Financial Fraud: An Application of Detection Controlled Estimation. Working paper, Wilfrid Laurier University.

Linck, J. S., J. M. Netter, and T. Yang. 2009. The effects and unintended consequences of the Sarbanes-

Oxley Act on the supply and demand for directors. Review of Financial Studies 22 (8): 3287–3328.

Lobo, G., and J. Zhou. 2006. Did conservatism in financial reporting increase after the Sarbanes-Oxley Act?

Initial evidence. Accounting Horizons 20 (1): 57–73.

Loughran, T., and J. R. Ritter. 1995. The new issues puzzle. Journal of Finance 50: 23–52.

Ludwig, J. 2006. Testimony before United States Senate Committee on the Judiciary. Washington, DC,

September 19.

Ludwig, J. 2010. The costs of crime. Criminology and Public Policy 9: 307–311.

Manzi, J. 2012. Uncontrolled: The Surprising Payoff of Trial-And-Error for Business, Politics, and Society. New York, NY: Basic Books.

Marosi, A., and N. Massoud. 2008. ‘‘You can enter but you cannot leave...’’: U.S. securities markets and

foreign firms. Journal of Finance 63: 2477–2506.

Michaels, A. 2003. After a year of U.S. corporate clean-up, William Donaldson calls for a return to risk-

taking. FinancialTimes.com (July 24).

Moore, S., and J. Shepherd. 2006. The cost of fear: Shadow pricing the intangible costs of crime. Applied Economics 38: 293–300.

Nagin, D., A. Piquero, E. Scott, and L. Steinberg. 2006. Public preferences for rehabilitation versus

incarceration of juvenile offenders: Evidence from a contingent valuation survey. Criminal Public Policy 5: 627–652.

Owens, E. L., and J. Wu. 2012. Window Dressing of Short-Term Borrowings. Working paper, Simon

School, No. FR 11-15. Available at: http://ssrn.com/abstract¼1811110 Picard, I. 2013. Trustee’s Ninth Interim Report for the Period Ended March 31, 2012. In Re Bernard L.

Madoff, United States Bankruptcy Court, Southern District of New York (April 30).

Piotroski, J. D., and S. Srinivasan. 2008. Regulation and bonding: The Sarbanes-Oxley Act and the flow of

international listings. Journal of Accounting Research 46 (2): 383–425.

Plumlee, M., and T. Lombardi Yohn. 2010. An analysis of underlying causes attributed to restatements.

Accounting Horizons 24 (1): 41–64.

Posner, E., and G. Weyl. 2013. Benefit-cost analysis for financial regulation. American Economic Review 103: 1–5.

Povel, P., R. Singh, and A. Winton. 2007. Booms, busts, and fraud. Review of Financial Studies 20: 1219– 1254.

Protiviti. 2012. 2012 Sarbanes-Oxley Compliance Survey. (May 2012). Available at: http://www.protiviti. com/en-US/Documents/Surveys/2012-SOX-Compliance-Survey-Protiviti.pdf

Qian, J., P. E. Strahan, and J. L. Zhu. 2009. The Economic Benefits of the Sarbanes-Oxley Act? Evidence from a Natural Experiment. Working paper, Boston University. Available at: http://fic.wharton. upenn.edu/fic/papers/09/0941.pdf

Raghunandan, K., and D. V. Rama. 2006. SOX Section 404 material weakness disclosures and audit fees.

Auditing: A Journal of Practice & Theory 25 (1): 99–114.

Reifschneider, D., R. Tetlow, and J. Williams. 1999. Aggregate disturbances, monetary policy, and the

macroeconomy: The FRB/U.S. perspective. Federal Reserve Bulletin (January): 1–19.

Rice, S., and D. Weber. 2012. How effective is internal control reporting under SOX 404? Determinants of

the (non-)disclosure of existing material weaknesses. Journal of Accounting Research 50 (3): 811– 843.

Rice, S., D. Weber, and B. Wu. 2013. Does SOX 404 have teeth? Consequences of the failure to report

existing internal control weaknesses. Available at: http://ssrn.com/abstract¼2239965 Roe, M. 2009. Delaware and Washington as corporate lawmakers. Delaware Journal of Corporate Law 34:

1–33.

Romano, R. 2005. The Sarbanes-Oxley Act and the making of quack corporate governance. Yale Law Journal 114 (7): 1521–1611.

SOX after Ten Years: A Multidisciplinary Review 669

Accounting Horizons September 2014

Ryan, E. M., and L. E. Simmons. 2012. Securities class action settlements 2012 review and analysis.

Cornerstone Research.

Sabol, T., H. C. West, and M. Cooper. 2009. Prisoners in 2008. Bureau of Justice Statistics Bulletin. Available at: http://www.bjs.gov/content/pub/pdf/p08.pdf

Scholz, S. 2008. The Changing Nature and Consequences of Public Company Financial Restatements: 1997–2006. Washington, DC: U.S. Department of the Treasury. Available at: http://www.treasury. gov/press-center/press-releases/documents/financialrestatements_1997_2006.pdf

Scott, H. S. 2007. Sarbanes-Oxley 404: Will the SEC’s and PCAOB’s New Standards Lower Compliance Costs for Small Companies? Hearing Before the House Committee on Small Bus., 110th Congress 2007. Washington, DC: Government Printing Office.

Securities and Exchange Commission (SEC). 2004. Management’s Report on Internal Control Over Financial Reporting and Certification of Disclosure in Exchange Act Periodic Reports, 69 Fed. Reg.

9,722. Washington, DC: SEC.

Securities and Exchange Commission (SEC). 2009. Study and Recommendations on Sections 404(B) of the

Sarbanes-Oxley Act of 2002 for Issuers with Public Float Between $75 and $250 Million. Washington, DC: SEC.

Shavell, S. M. 1986. The judgment proof problem. International Review of Law and Economics 6: 45.

Shleifer, A., and R. Vishny. 2003. Stock market driven acquisitions. Journal of Financial Economics 70:

295–311.

Sikka, P. 2009. Financial crisis and the silence of the auditors. Accounting, Organizations and Society 34: 868–873.

Srinivasan, S., A. Wahid, and G. Yu. 2014. Admitting mistakes: Home country effect on the reliability of

restatement reporting. The Accounting Review (forthcoming).

Szewczyk, S. H., and G. P. Tsetsekos. 1992. State intervention in the market for corporate control: The case

of Pennsylvania Senate Bill 1310. Journal of Financial Economics 31: 3–23.

Thaler, R. 1978. A note on the value of crime control: Evidence from the property market. Journal of Urban Economics 5 (1): 137–145.

Valukas, A. 2010. Report. In Re Lehman Brothers Holdings, Inc., U.S. Bankruptcy Court, Southern District of New York, Volume I, March 11.

Van de Poel, K., and A. Vanstraelen. 2011. Management reporting on internal control and earnings quality:

Insights from a ‘‘comply-or-explain’’ internal control. Auditing: A Journal of Practice & Theory 30

(3): 181–209.

Velikonja, U. 2013. The costs of securities fraud. William & Mary Law Review 54: 1887–1957.

Viscusi, K. 2000. The value of life in legal context: Survey and critique. American Law and Economics Review 2: 195–222.

Vogel, G. 2009. Despite initial fears to the contrary, it appears that Sarbanes-Oxley gave private litigants a

‘‘dull sword’’ when it comes to piercing the corporate veil. Fordham Journal of Corporate and

Financial Law 14: 415–442.

Wang, T., A. Winton, and X. Yu. 2010. Corporate fraud and business conditions: Evidence from IPOs.

Journal of Finance 65: 2255–2292.

Wang, T. 2013. Corporate securities fraud: Insights from a new empirical framework. Journal of Law,

Economics, & Organization 29 (3): 535–568.

Weiss, A., and G. Birkner. 2008. Charities, day schools hard hit by Madoff scandal. Forward.com

(December 17).

Whalen, D., C. McCoy, and P. Cross. 2013. Audit fees and non-audit fees: An eleven-year trend. Audit

Analytics.

Woo, C. 2011. United States securities regulation and foreign private issuers: Lessons from the Sarbanes-

Oxley Act. American Business Law Journal 48: 119–176.

Yeh, Y., H. Chung, and C.-L. Liu. 2011. Committee independence and financial institution performance

during the 2007–08 credit crunch: Evidence from a multi-country study. Corporate Governance: An

International Review 19 (5): 437–458.

670 Coates and Srinivasan

Accounting Horizons September 2014

Zakolyukina, A. 2013. Measuring Intentional Manipulation: A Structural Approach. Working paper, Booth School of Business.

Zhang, I. 2007. Economic consequences of the Sarbanes Oxley Act of 2002. Journal of Accounting and Economics 44: 74–115.

SOX after Ten Years: A Multidisciplinary Review 671

Accounting Horizons September 2014