SEC Independence Concepts

IN SEARCH OF SOLUTIONS
By Robert K. Elliott and
Peter D. Jacobson

In Brief

Sorting Out What Really Matters

In making a presentation to the new Independence Standards Board, the staff of the SEC proposed its own test of auditor independence.

The authors address the concepts in that test and explore other concepts of independence. Among the concepts addressed are the following:

• Mutual interests--This principle covers some necessary ground but does not work as an all covering principle.

• Conflicting interests--The concept is useful when it is tied in some direct way to the quality of the audit but is not useful when tied to the client as in the SEC staff's independence test.

• Appearance of independence--Appearance alone, apart from the fact of independence, should not be the basis of regulatory requirements.

• Objectivity--A new model is proposed that would separate independence from objectivity. In this model, independence is not a mental attitude or state of mind.

• Knowledge of all the related facts--Knowledge of all the related facts and circumstances should include an understanding of the regime designed to ensure independence in fact and an awareness of the interests weighing in favor of audit independence.

With the new Independence Standards Board (ISB) up and running, SEC concepts will be front and center, at least initially. The ISB operates in the sunshine; it has begun an examination of independence concepts; and until new standards are created, the SEC's independence rules are the ISB's standards. Those rules will be the point of departure for any change in requirements by the ISB, and SEC concepts must be considered in determining any new conceptual framework.

The SEC staff made a presentation to the ISB on October 20, 1997, and distributed a handout. Its text puts the staff's concepts of independence in the form of this test of auditors' independence:

Whether a reasonable investor, knowing all the relevant facts and circumstances, would perceive the auditor as

1) having neither mutual nor conflicting interests with its audit client, and

(2) exercising objective and impartial judgment on all issues brought to the auditor's attention.

This statement is the most current crystallization of the staff's concepts. Its ideas are the subject of this article, which examines the meaning of mutual and conflicting interests, then the treatment of appearance of independence as a separate, sufficient cause for specific regulation. Next the SEC staff's use of the concept of objectivity is explored, and, finally, the concept, "knowing all the relevant facts and circumstances."

Mutual Interests

The SEC staff holds that an accountant should have "neither mutual nor conflicting interests with its audit client." The concept is stated in terms of the reasonable investor's perception of the auditor, but it must also be addressed as a prohibition of a relationship incompatible with independence. In past rulings on independence, "mutuality of interests" is simply treated as prohibited behavior.

The no-mutual-interests principle covers some necessary ground. It works to the degree that it identifies situations where the auditor's interests can create bias (e.g., financial interests in the company, taking on the role of management). However, it does not work as an all-covering principle.

The language and its usage, taken literally, mean that the interests of the auditor and the client should never be the same: They should never have shared interests. The words imply that if the auditor is interested in the truth of the financial statements, as should always be the case, management is interested in their untruth. Thus, the no-mutual-interests principle implies that management is bent on fraud.

Prejudging management negatively is not the act of an independent, objective observer. An independent, objective observer ideally has no prejudices other than a partiality toward the truth. The independent auditor has obligations to search for evidence, including evidence of fraud, but should not in the interests of objectivity assume the outcome of the search.

The no-mutual-interests principle runs hard on the desired relationship between the auditor and the shareholders. If the auditor has no mutual interests with management and management has the same interests as shareholders (e.g., increasing shareholder value), the auditor has no interests in common with shareholders.

Enlightened management is likely to share the auditor's and shareholders' interest in fair disclosure, especially when the company is profitable and growing. Management needs good relations with shareholders and with the investment community because good relations reduce the cost of capital. Managements thus have a business interest in fair disclosure that can make for mutual interests between the auditor and the client but is compatible with audit independence. Shareholder relations departments, meetings with analysts, and voluntary disclosures all suggest that many management teams understand the relationship between disclosure and the availability and cost of capital.

Conflicting Interests

Avoidance of conflicting interests is a helpful independence concept when it is tied in some direct way to the quality of the audit (for example, the auditor should have no interest that conflicts with the quality of the audit), or, better, tied directly to the auditor's interest in the quality of the information that is the subject of the audit. But it is not useful as an independence concept when tied to the client as in the SEC staff's independence test. If management were engaged in fraud, following the no-conflicting-interests principle would render the auditor incapable of being opposed to the fraud. The fraudster managers would have an interest in the success of the fraud with which the auditor could not be in conflict.

Before going on, a brief comment on relationship between the two elements of the dual stricture: neither mutual nor conflicting interests with the client. Full compliance with the letter of the dual stricture would leave the auditor with no interests related to the client. That would exclude the legitimate situation--the audit itself--that the requirements are supposed to govern. But full compliance would be impossible. The audit cannot exist without generating either common or conflicting interests between the client and the auditor. For example, the auditor would have an interest in conducting the audit efficiently that the client would most likely share.

Appearance of Independence

The SEC-staff test of independence quoted above is based on perceptions. It is therefore consistent with the staff's approach to independence rulings, which treat both the fact and the appearance of independence as separately and equally necessary. Each, under the staff's approach, is by itself a sufficient cause for specific regulation. The AICPA's code of ethics takes the same position. In the SEC's case, no other concept has been more prominent than appearance in complaints about auditors' independence in the past few years. The AICPA, in its recent white paper, challenged the notion that appearances have been validly used as a basis for independence decisions, but accepted that research on appearances, properly done, could be a valid basis for independence requirements. However, appearances alone--apart from the fact of independence--should not be the basis of regulatory requirements.

The appearance of independence is fundamentally different from the fact of independence. The two are not obvious partners for independence policy, as these points show and as is illustrated in Exhibit 1.

Investors do not suffer damage from specific engagements when, because of appearances, some mistake an independent auditor for one who is not independent. In such cases, the investors do not rely on that auditor's report. The party damaged in that situation is the auditor, who is independent but losing credibility in the marketplace and will eventually be out of business. Corporate clients will not want the auditor whose reputation represents a higher cost of capital. It is true that there might be a negative effect on share price caused by the misimpressions, but, as argued at the end of this section, that is an acceptable risk.
• Even if, despite the appearance of lapsed independence, investors rely on the auditor's report, they would not be in any danger, provided, again, the auditor is independent in fact. Compromised appearance of independence does not affect objectivity or the performance of an engagement. The auditor's report would be valid and reliable.

• Lapses in the fact of independence differ on both counts. The public relies because they believe the auditor is independent, and the public is vulnerable to injury because lapsed independence in fact can mean defective audit performance.

The practical upshot of these points is that regulation--i.e., independence requirements--should be based on the fact of independence, not on supposed appearances. If effective requirements are based on the fact of independence, the public will be fully protected. And if this is true, the efforts to delve into appearances and to base regulation on suppositions about appearances, or even research findings about them, is unjustifiable restraint of trade. The regulator would be prohibiting activity in the conduct of commerce on the basis of its appearance when the appearance threatens injury to no party other than the one pursuing the activity.

The rationale underlying support for separate treatment of appearances is the need to assure confidence in the disclosure system and capital markets. At first glance this seems persuasive. Confidence in the marketplace is incontestably important. The auditor's contribution to that confidence is also incontestably important. But it does not follow that the best way to preserve and protect such confidence is through rules about the appearances of auditors who are independent in fact. And it does not follow that restraint of trade is justified in the hope of achieving improved confidence in the markets through regulated appearances.

The fact of independence is a valid basis for confidence in the capital markets. The appearance of independence alone (i.e., apart from any relationship to the fact of independence) is not. It is true that appearances can lead to conclusions about the fact of independence, but in that instance it is only the fact of independence that is at issue. There is no separate justification for regulation on the basis of appearance divorced from the fact of independence. The section below on knowledge of all the relevant facts gives more on this point.

Investors and creditors are well aware of the difference between appearances and facts. They know that fraud is a confidence game and that hyped marketing is a condition of modern economies. What they need, and presumably know they need, is reliable information for their decisions. The reliability of financial information is improved by quality auditing, which depends on independence in fact. Reliability benefits not a jot from the auditor's appearance. A rational investor should be wary of an auditor who appears independent but in fact is not, but should have no fear of an auditor who is known to be independent in fact, but whose appearance has nevertheless been questioned. There is no reason to believe the investment community, to the extent it actually evaluates accountants' independence, evaluates it by a standard other than the fact of independence.

Public appreciation of a reliable regime to ensure the fact of independence should therefore be the surest way to boost the appearance of independence. It is probably the only basis on which appearance of independence can be maintained and improved. People and groups are easily defamed in our society. The press can be merciless, and the public is often eager to believe the worst. In this environment, no campaign to improve the appearance of independence not based on the fact of independence can work. Such a campaign, if pursued in earnest, not with mere prohibitions based on supposed appearances, would turn into a public case for the fact of independence even if it started out based on appearances. A rational, tough-minded investor should care about the fact of independence, because it affects his or her interests. Appearances, apart from the fact of independence, do not.

Even assuming one could manage a profession's credibility apart from the quality of its performance, the appropriate parties for the task would be experienced, full-time advertising and public relations experts, not regulators. The exclusive focus on appearances is about image apart from substance. It is the province of those sometimes called spinmeisters.

The current regulatory trend is to trust the marketplace and its decisions except when there is a clear need for regulation. Regulating the appearance of independence goes against that trend. It assumes consumers cannot make decisions in their interests, whether they are companies firing auditors to improve or maintain their cost of capital or investors using audit reports. Regulating the fact of independence, on the other hand, protects consumers' ability to make decisions in their interests. It helps lower the risk that investors make ownership and credit decisions based on faulty information. The marketplace can adjust to defective appearance of independence, which by definition is something it knows about, but it cannot adjust to instances of lapsed independence in fact that it does not know about.

To a great extent, the question is whether auditing should be conceived as a marketplace activity. We know that Congress, in passing the securities laws, rejected the notion of government auditors in favor of private-sector auditors. Auditors were to operate in the marketplace. If auditing is a marketplace activity, it will be most productive when qualified auditors are permitted to perform their work in that marketplace.

Regulating appearances apart from the fact of independence runs implementation risks. The way would be open for regulators' personal likes and dislikes to play a more prominent role in the regulatory process, because appearances are subjective and the subjectivity problem redoubles when assessing how things appear to others. Opinion research would not be without heavy risks of arbitrariness. Findings must be interpreted, allowances made for imperfections in technique, and tolerable and intolerable levels of group feeling defined. Doubts about accepted truths are often introduced by questions, and the way in which they are asked can lead to different answers on the same issue. Opinions change. If a research finding on appearances led to regulation prohibiting some behavior, the public pulse would have to be taken periodically to adjust the regulation, following adoption with repeal when the wind of opinion shifted. Given the fact that no external party is injured by the appearance of an auditor independent in fact, can such an enterprise be cost-effective?

The greatest implementation risk is that the public focus on appearances apart from the fact of independence will create the problem it is trying to forestall. It would be bad publicity. The process would throw a distrustful light on auditing even assuming flawless audit competence, unblemished integrity, and independence in fact, because regulators determining what is permissible by appearances would be publicly questioning the fitness of auditors. This makes it more likely that the approach would damage rather than improve confidence in the markets.

We return to the issue of the potential negative effect on a company's share price caused by investors' reactions to the impaired appearances of auditors who are independent in fact. This is a risk from not regulating appearances apart from the issue of independence in fact. However, it is a risk that should be accepted for these reasons. First, any such price effect should be short-term and righted by the truth about the company, which would be on record and persist because the auditor is independent in fact. Second, attempting to prevent such misimpressions by regulating appearances apart from the issue of independence in fact runs the implementation risks just described, including the risk of diminished investor confidence in the markets that could have price effects. Third, as already noted, regulating appearances is inconsistent with relying on the marketplace except when there is a clear need for regulation. The audit marketplace can adjust to impaired appearance of independence, but cannot adjust to impaired independence in fact that is not indicated by appearances. Fourth, effective regulation of independence in fact should minimize such instances, because it is the real basis of confidence in auditors. Indeed, it is doubtful that an instance of diminished share price from investors' concerns about appearance alone (i.e., apart from the fact of independence) can be cited. They are likely to be very rare.

Can It Be Explained?

It is difficult to explain the regulatory complaints about appearance of independence in recent years. The SEC's governing rule on independence (Rule 2-01, on the qualifications of accountants) does not mention appearances. The SEC's later focus on appearances as a separate, sufficient cause for regulation is arguably inconsistent with Rule 2-01. During the recent period of regulatory complaints about appearances, the SEC has had full control of independence rules, which incorporate restrictions based on the notion of protecting appearances. There is good reason to believe that firms have followed those rules to the letter. No wonder the Kirk report concluded in 1994 that there was no reason for additional rules or regulations. In fact, the SEC staff had reached and reported the same conclusion earlier that year. Yet some SEC leaders complain about the situation that exists with their rules in place.

It is possible that statements about the appearance of independence originate in a desire to avoid the damning accusation that an activity or relationship makes an auditor not independent in fact. To say an auditor is not in fact independent can have the effect of defrocking. It deprives the auditor of a professional credential and can apply a lasting taint. Saying appearances are not what they should be is a less degrading alternative. It is likely that the notion of appearance of impropriety plays a role in nonaccounting ethical situations (e.g., politics) at least in part for this reason.

It is extremely difficult to argue successfully that independence in fact suffers from the nonaudit services now performed, so that line of opposition to nonaudit services has not seemed open. The argument runs into these hard facts:

• There is no known case of a nonaudit engagement damaging audit quality, despite intensive study by respected committees and individuals, including scholars, over many years. Indeed, the history of performing nonaudit services goes back to the origins of the profession in this country. There never was a time when auditing firms did not perform them.

• The independence rules governing the provision of nonaudit services to audit clients were written by the SEC, and they stand in addition to rules devised by the AICPA.

• Independence is studied during peer review (because independence is a quality control); the Public Oversight Board virtually participates in the peer review process, but most certainly examines it in detail; the SEC also exercises oversight and annually for over a decade has said that peer review contributes to audit quality.

• Although, as many have pointed out, the fee from nonaudit services creates an interest in retaining the client that could diminish independence, the net financial effect of nonaudit services on the firm is leverage for independence. The firm with nonaudit services adding to revenue totals for its practice is less dependent on the fee from any single client and therefore in a better position to stand up to the client in the event that differences occur.

• Massive economic incentives weigh on the side of retaining independence-- the firm's reputation and the threat of litigation. Audit failures are high profile incidents, high enough in profile to damage a firm's ability to sell any of the services in its portfolio, whereas a reputation for objectivity and independence from effective auditing helps sell other services. The economic incentives affecting independence are analyzed in detail in the AICPA white paper.

The profession's stance has contributed to the SEC's focus on appearances. For over a generation the profession's authoritative literature has taken the position that the auditor should be independent in appearance as well as in fact (see SAS 1). SEC regulations on independence cite the profession's literature in making the case for appearances.

We can speculate on one other part of the explanation. The profession has been changing. It must grow its nonaudit services to be economically successful. Some of those recalling the years when auditing was an absolute monarch in accounting firms find the change wrenching. However, so long as independence and audit quality are maintained, the change serves firms' audit segments. It is inconsistent to want audit services maintained in their excellence and also to support measures that would weaken them economically.

Objectivity

The SEC staff's test for auditor independence includes whether a reasonable investor would perceive the auditor to be "impartial and objective" in judgment. Since the paired terms are virtual synonyms, only the usage of "objective" will be questioned.

The SEC's concept means the auditor who is independent is simultaneously and necessarily objective. This is not the most useful way to conceive of the relationship between independence and objectivity. For one thing, objectivity is a variable. Some people have more of it than others. They may be less wedded to patterns of thought and assumptions, for instance, and therefore be able to view new perceptions less prejudicially. Thus, in the best of circumstances a given auditor can bring no more than the measure of objectivity he or she has. Independence cannot guarantee a perfectly objective audit. Apart from that, it is more useful for purposes of a conceptual framework to distinguish between objectivity and independence and to articulate the relationship between them and between them and integrity.

Independence pertains to interests (e.g., financial or psychological) that might interfere with objectivity ("might," because the interests also might not have that effect). Objectivity is an intellectual quality; integrity is a moral quality. Because independence and integrity can function separately, they protect the public more than if objectivity could be undermined only by the interests we think of as compromising independence.

In the case of the audit, with its moral obligations, when the auditor's objectivity is damaged by lack of independence, his or her integrity is also damaged. However, as addressed in the next paragraphs, integrity can be unaffected by interests and therefore protect objectivity.

Exhibit 2 shows the relationships among the ideas and how they interact to influence the quality of an audit. The purpose of independence is to contribute to a quality audit, and the purpose of a quality audit is to improve the reliability of decision-making information and thereby contribute to effective capital markets.

For a CPA to render a quality audit, he or she must have objectivity (which limits bias) and competence (so the right facts and conclusions can be elicited and drawn). The auditor's objectivity--freedom from bias--would be at its maximum if he or she had perfect integrity. We all know people who are so honest that they would sooner impair their self-interest than tell a lie. Even if the auditor had interests that would undermine objectivity in a person of lesser integrity (e.g., shares in the audited company), they would not impair objectivity in a person with perfect integrity. In this way, integrity can help assure objectivity, giving the public protection in addition to the absence of interests that could cause bias (independence).

Thus there is some redundancy in the model: Either perfect integrity or perfect independence would suffice to maximize the auditor's objectivity. But "good enough" integrity plus "good enough" independence, in tandem, will eliminate all but the worst cases of lack of
objectivity.

This model is closer to the real world than the notion of absolute independence leading to absolute objectivity. The audit fee is an interest that theoretically could undermine objectivity. We accept it because we have good reason to believe it does not prevent objective audits and because no other approach has greater claims to cost-effectiveness.

The model explicitly separates independence from objectivity. As stated above, independence is a matter of interests that might cause bias, whereas objectivity is an intellectual quality (degree of absence of bias). Thus, independence in this model is not a mental attitude or a state of mind, terms that have been frequently used to describe independence. When a regulator evaluates auditors' independence, he or she is weighing whether an interest or set of interests represents an unacceptable risk of material bias with respect to the reliability of the financial-statement information. This approach is quite different from treating independence as either the same as objectivity or tantamount to being the same.

The model is consistent with the report on assurance independence prepared by the Special Committee on Assurance Services. A fuller explanation of some of the ideas can be found there and in our earlier article on audit independence that appeared in the March 1992 issue of The CPA Journal.

In addition to auditor's independence and integrity, the public interest in quality audits is protected in other ways. For example, it is protected by quality controls, peer reviews, and SEC reviews of filings. Independence is best viewed in the full context of what makes for a quality audit.

Knowledge of All the Relevant Facts

The SEC's conceptual measure of audit independence is how "a reasonable investor, knowing all the relevant facts and circumstances, would perceive the auditor." This could derive from the legal notion of the reasonable person or from a version of democracy where the opinions of actual investors determine what is permissible for purposes of independence. Item (c) of Rule 2-01, the governing rule for the commission's independence program, says the SEC will make determinations of independence considering all the relevant circumstances. Item (c) is consistent with the legal notion of the reasonable person.

The democratic alternative, with its hints of votes, tallied opinions, or some other approach to seeking a majority or better, is not practicable. Audits cannot wait for polling to rule on an audit team with some new particulars of interests, and it isn't obvious who would be polled in such instances (the shareholders who have already approved the audit firm by proxy, some version of potential investors in the company, or some surrogate for the entire array of "users"). Even for standards, the democratic notion is not practicable. Neither the SEC nor the ISB can expect to get significant participation when the FASB, whose standards directly affect what corporations do in financial reporting and what analysts do when evaluating companies, has always had limited, disappointing participation by investors. Division of labor and delegation of responsibility are typical of our society and its institutions. If every person were responsible to participate in decision making on all matters that affected him or her, from traffic rules to the purity of soap, little would get done. This brings one back to the legal notion of the reasonable person concept.

In the context of independence, it means arriving at regulatory decisions using as a measure the hypothetical perceptions of the reasonable, fully informed investor. The hypothetical investor's perceptions become the test of the no-mutual-interests and the no-conflicting-interests principles. This is consistent with due process and obtaining input from all interested parties who choose to contribute, but the final decision by the regulator would be made using the reasonable-investor criterion. The threshold question is what it means in applying the reasonable-investor criterion to have knowledge of all the relevant facts and circumstances.

Knowledge of all the relevant facts and circumstances should include an understanding of the regime designed to ensure independence in fact and an awareness of the interests weighing in favor of audit independence. This applies to participants in research as well as to the hypothetical reasonable investor. No party should be asked to evaluate the fact or the appearance of independence unless aware of the rules and the safeguards to produce independent, quality audits.

Both the hypothetical reasonable investor and any research subjects should know, for example, that there are detailed independence rules, the rules were formulated by the SEC, there are powerful reasons to believe the rules are followed (peer review, SEC and POB oversight, economic interests, etc.), no one but the auditor is injured by appearances alone, and the performance of nonaudit services--the cause of so much concern about supposed appearances--is not known to have damaged any audits. They should know about the policies and procedures to ensure audit quality--the educational, experience, and continuing-education requirements for CPAs, for example, and the range of quality controls. Questions about appearances are likely to get different responses from those who know and those who don't know that there is good reason to believe auditors are independent in fact, their competence has been assessed, their performance is tested in the interests of audit quality, and the penalties for lapsed independence are staggering.

Arguably, awareness of all the facts and circumstances would lead the hypothetical reasonable investor or a research subject to judge the fact of independence, not the appearance. The two would be substantively the same. This means judging appearances alone and making rules on that basis is inconsistent with being aware of all the relevant facts and circumstances and applying them. All that would be judged is the sliver of the facts (assuming, for now, they are not rumors or erroneous allegations) that represents the appearance. An incomplete set of facts creates a distorted picture.

Selective disclosure of this sort may have occurred when ASR 250's requirements were in force, starting in 1978 and running until it was rescinded a few years later. The required disclosures included the types of nonaudit services performed and relationships between the fees for the nonaudit services and the audit fee. The disclosures were a sliver of the facts about the auditor's relationship to the client and its context, a sliver of the facts and circumstances by which a hypothetical, fully informed, reasonable investor would judge independence. There was no reason to suppose that those who might have read the disclosures otherwise had access to all the relevant facts and circumstances or were familiar with them. Few nonauditors, even savvy investors, know the systems and interests that help ensure independence and audit quality or the record on the relationship of nonaudit services to audit quality. Selective disclosure of the sort the SEC required by ASR 250 should not be expected to have a fair influence on readers. It may have no influence, but it is unlikely, if influential, to have a fair influence. Fair or not, the approach would violate the notion that judgments on independence should be made in awareness of all the relevant facts and circumstances.

A premise underlying the notion that investors could use such information is that they spend time evaluating auditors' independence. As already noted, there is no reason to believe the investment community would evaluate accountants' independence by a standard other than the fact of independence. But it is not likely that investors actively evaluate even the fact of independence. It is hard to imagine busy, intelligent professional investors regularly attempting to learn the interests of audit team members that could create bias. They are far more likely to rely on the author of the rules in place, the institutions responsible to ensure compliance, and awareness of the auditor's historic and ongoing obligations, using these to make a reasonable assumption about independence. In the end, independence is accepted by investors, not determined by them from specific evidence. If this interpretation is true, undermining their acceptance by selective disclosures or by otherwise questioning in public the fitness of qualified auditors independent in fact would needlessly damage confidence in the marketplace.

Robert K. Elliott, CPA, is a partner, and Peter D. Jacobson, PhD, a policy analyst, in the national office of KPMG Peat Marwick LLP. Mr. Elliott served on a task force assisting in the development of the AICPA's white paper on auditor independence. However, this article represents only the views of the two authors. It is not intended to communicate views of the AICPA or other parties.