Behavioral evidence on the effects of principles- and rules-based standards. (Commentary).

INTRODUCTION

The Sarbanes-Oxley Act requires the SEC to study the feasibility of shifting to a more "principles-based" financial reporting system, and the FASB has proposed changes designed to create a more principles-based approach to standard setting (FASB 2002a). This commentary

The "Incremental" Perspective

Because U.S. accounting standards typically are written to operationalize the FASB's underlying conceptual framework, they are based on principles. The standards also provide guidance as to correct accounting or disclosure treatment, so they include rules. As a consequence, an immediate challenge in writing this commentary is to define exactly what it means for a standard to be "principles-based" or "rules-based." I restate the "principlesbased vs. rules-based standards" question in a way that is more conducive to extracting insights from the existing research literature. Assuming that the FASB continues to base standards on the framework of principles articulated in the FASB's concepts statements, all standards can be viewed as principles-based, and the issue is the incremental effect on behavior when standards include relatively more elaborate rules. Therefore, in this commentary I discuss standards in terms of being more or less rules-based, acknowledging that less rules-based standards must rely more on p rinciples to guide behavior. I define "rules" broadly to include specific criteria, "bright line" thresholds, examples, scope restrictions, exceptions, subsequent precedents, implementation guidance, etc. A "standard" is the total body of principles and rules that apply to a given accounting issue.

This incremental perspective is useful in at least two ways. First, it focuses the review on the incremental effects of increasing the number of rules in a standard. I argue that adding rules affects the precision and complexity of an accounting standard, and consider the implications of studies that examine the effects of precision and complexity on the behavior of participants in the financial reporting process.

Second, the incremental perspective may provide insight about the effect of varying the extent to which different standards appear to be principles-based or rules-based. Some accounting topics involve complex transactions with predictable characteristics, so it is possible to write standards that contain very specific thresholds. For example, leases typically involve physical assets and a contractually defined sequence of cash flows, and SFAS No. 13 (FASB 1976) uses numerical thresholds to specify criteria regarding the proportion of lease life and the amount of fair value. These sorts of accounting topics are amenable to rules, but do not necessarily require them, so standard setters must choose the extent to which the standard is to be rules-based. Other accounting topics have less predictable characteristics and more inherent judgment. For example, SFAS No. 5 (FASB 1975) involves estimates of probability and amounts that are not defined contractually. Because these topics are less amenable to rules or brig ht line thresholds, the standards that govern them are more principles-based by default. Finally, including precedents and implementation guidance in the definition of rules highlights that the extent to which a standard appears rules-based changes over time, as a particular standard accretes implementation guidance, interpretations, and technical rulings. One reason why relatively younger standard-setting regimes like the IAS appear more principles-based is that they have not had as much time to accrete rules.

I review the incremental effects of rule precision and complexity on performance with respect to two important functions of financial accounting standards: communication and constraint. Communication refers to the role standards play in conveying GAAP to practitioners and facilitating a shared understanding of the meaning of financial reports. Assuming incentives to report accurately, how does the extent to which standards are rules-based influence their ability to communicate? Constraint refers to the role of standards in discouraging biased communication by serving as the benchmark for assessing and defending the accuracy of communications. Assuming incentives to report inaccurately, how does the extent to which standards are rules-based constrain biased communication? Following separate discussions of research on communication and constraint, I discuss the trade-offs between them.

Focus on Experimental and Survey Research

I review research from financial accounting, auditing, and tax, recognizing that although these areas differ in important ways, basic behavioral responses generalize across areas. I focus on evidence provided by experimental and survey studies. To my knowledge, few archival studies examine whether decision makers alter their behavior depending on the precision and complexity of relevant accounting standards. (1) Also, because this review focuses on observed behavior, I exclude analytical studies in law and accounting that predict and/or model how behavior varies with rule precision and/or complexity. (2)

The relative strengths and weaknesses of research approaches influence their ability to address issues relevant to accounting policy. Good experiments abstract from practice and randomly assign participants to alternative treatments (Kachelmeier and King 2002; Libby et al. 2002). This approach provides much control and enables strong causal inferences from directional changes in observed behavior, but also requires care when generalizing to practice. Good surveys elicit detailed information from a broad range of sophisticated practitioners (Gibbins 2001). This approach provides rich insight into the practice setting but involves self-reported data of uncertain validity and generality. Given these methodological strengths and weaknesses, I present research findings in terms of general behavioral responses to particular characteristics, such as, "when the precision of a standard increases, people will tend to do X more frequently," rather than attempting to predict the exact levels of behavior that occur under a particular type of standard.

COMMUNICATING GAAP

I consider two ways in which standards can increase their communication accuracy: (1) standards can include precise, "bright line" thresholds, and (2) standards can use a larger number of rules.

Quantitative and Qualitative Thresholds

Standards often use "bright line" criteria to distinguish between broad classes of transactions that have different accounting treatments. (3) For example, SFAS No. 13 requires capital-lease treatment when the lease term is equal to 75 percent or more of the estimated economic life of the property. As an alternative, SFAS No. 13 could have required capital-lease treatment if the lease term constituted most of the estimated economic life of the property, but the meaning of "most" is less clear than the meaning of "75 percent." Given that practitioners can calculate the amounts that are compared to the precise threshold, bright line descriptions of quantity communicate accurately, and appear throughout the financial accounting standards, as illustrated by FASB Concepts Statement No. 2 (1980, Table 1).

Less precise expressions are often used to communicate other types of thresholds. For example, SFAS No. 5 (FASB 1975) requires accruing a loss contingency when the loss can be reasonably estimated and it is probable that a liability has been incurred or an asset impaired. One concern about the use of such imprecise expressions is that practitioners will interpret them differently, leading to low consensus among practitioners and potential lack of comparability between financial reports. Several studies that elicit numerical interpretations of the SFAS No. 5 probability expressions from experienced auditors support this concern. For example, Amer et al. (1994) find relatively high variance in audit managers' interpretations of probability expressions, suggesting that these expressions sometimes are interpreted differently than standard setters intended when applied in a given context. Amer et al. (1995) also provide evidence that audit managers' interpretations of "probable" vary predictably between contexts i n which SFAS No. 5 is applied, in a manner that was not intended by the FASH members who promulgated SFAS No. 5.

High variance in probability-threshold interpretations between people and between contexts could encourage switching to bright line thresholds. For example, SPAS No. 109 (FASB 1992) defines "more likely than not" as "more than 50 percent probability" (para. 17c). However, the psychology and accounting literatures indicate that such a change might not be effective in reducing variance in interpretations, because people often struggle when estimating and interpreting numerical probabilities. Wallsten et al. (1993) summarize psychological research that uses inexperienced participants to perform generic tasks, and conclude that there is generally little difference in outcomes of judgments made using numerical and verbal probabilities. In accounting, the primary studies are set in the context of auditors' evaluations of components of the audit risk model. Participants in these studies provide risk ratings using their firms' scale of probability phrases, such as "low" risk or "moderate" risk. Researchers translate these ratings into numerical probabilities, using either participants' numerical definitions of the phrases or definitions provided by the auditing firm. Stone and Dilla (1994) use these techniques to provide evidence that consensus in auditors' risk judgment is higher when assessments are based on numerical probabilities, while Reimers et al. (1993) and Dilla and Stone (1997) report that consensus is higher when risk judgments are based on probability phrases. In general, this research suggests that replacing probability phrases with numerical thresholds does not materially enhance the accuracy with which standards communicate.

Number of Rules and Dimensions of Complexity

Another way to enhance the communication accuracy of a standard is to increase its precision by including more rules. Standards can explicitly allow or disallow individual accounting treatments or provide examples and detailed implementation guidance. For example, SPAS No. 125 (FASB 1996) forbids de-recognition of liabilities through insubstance defeasance, and provides numerous illustrations of appropriate implementation of the standard. These sorts of rules can be included in the original standard, but also may accrete over time as precedents and guidance accumulates. The problem with this approach to increasing precision is that it can increase the complexity of the standard, thereby creating communication problems that offset the communication benefits provided by increased precision.

Practitioners often complain that voluminous rules create such a "standards overload" that very few practitioners are able to accumulate and absorb the complex information that standard setters are trying to communicate (Shaw 1995; Beresford 1999). Considerable research examines how task complexity affects judgment; see Bonner (1994) and Tan et al. (2002) for reviews. Task complexity generally harms judgment accuracy and consistency by encouraging coping strategies that reduce mental processing, such as disregarding potentially important information and combining information in simplistic ways. According to Wood (1986), total task complexity is determined by task characteristics that affect "component complexity," "coordinative complexity," and "dynamic complexity." In the context of applying financial reporting standards, component complexity increases with the number of decisions to be made and the number of precedents and examples to be considered. Coordinative complexity increases when information must be combined in complex or unspecified ways when determining whether a standard is satisfied. Dynamic complexity increases when the requirements necessary to satisfy a standard shift over time.

Wood's (1986) definition indicates that additional rules may increase or decrease task complexity, depending on the circumstances. On the one hand, adding an exception or precedent presumably increases total task complexity, because component complexity increases with more rules, coordinative complexity increases when a new rule must be considered in light of existing rules, and dynamic complexity increases by changing the pattern of rules over time. On the other hand, adding an implementation guideline that sequences the decisions necessary to implement a standard, or adding an index that better relates existing rules and precedents, could lower total task complexity by reducing coordinative complexity more than the addition increases component and dynamic complexity.

Complexity may not reduce communication accuracy for all practitioners to the same extent. Rather, prior experimental research indicates that practitioners who possess more relevant knowledge, such as search strategies and indicators of information relevance, are better able to cope with complexity (Bonner and Lewis 1990; Libby and Tan 1994). For example, As are and McDaniel (1996), Tan and Kao (1999), and Tan et al. (2002) all provide evidence that auditors perform better on complex tasks when they have higher task-relevant knowledge and ability. Cloyd (1995) finds that tax professionals who are more knowledgeable about tax rules are better able to quickly locate relevant rules, and Spilker (1995) reports that experienced tax professionals are better than graduate tax students at identifying keywords for use in searching for relevant tax authority.

This finding that knowledge helps practitioners sift through relevant authority could explain some of the outcomes of auditor/client negotiations reported by Nelson et al. (2002) (hereafter, NET). NET surveyed 253 audit managers and partners, eliciting 515 descriptions of earnings-management attempts that included the transactions involved, the relevant accounting guidance, and the auditors' decisions to either require adjustment of the attempt or waive adjustment. The cell labeled "C" in Table 1 shows NET's evidence that auditors are relatively likely to require adjustment of their clients' aggressive accounting when the client has not structured transactions to comply with precise guidance as to appropriate treatment. (4) Auditors often explained that these situations occurred because the client appeared unaware of the relevant accounting rules. To the extent that auditors know more than their clients about the relevant standards, or can consult more knowledgeable resources, they are better able to cope wit h the complexity of the relevant accounting regulations and find guidance specific to client circumstances.

Another role of structure in professional standards may be to assist practitioners in drawing analogies. Standard setters recognize that, without precise rules, practitioners must reason by analogy, mapping relations between elements of standards or examples to their own decision problems (SAS No. 69, AICPA 1992, para. 9). Research in psychology and tax provides evidence that this mapping depends on decision makers seeing through surface features of a problem to identify the key relations that determine the structure of the analogy (Marchant et al. 1993). These studies suggest that even standards with relatively few rules could benefit from increased structure and carefully chosen examples that facilitate analogy development.

Communication: Summary

The research literature suggests that bright line thresholds can be used in some circumstances to communicate accurately. However, the more general way to increase the precision with which a standard communicates is to increase the amount of specified decision process, detailed implementation guidance, examples, precedents, and other rules that are in the standard. Although additional rules increase precision, they also increase various dimensions of complexity, unless they reduce coordinative complexity by increasing structure. Standard setters face a trade-off between including too few rules and creating a standard that communicates too vaguely and is interpreted inconsistently vs. including too many rules and creating a standard that becomes so complex that parts of it are applied incorrectly or missed entirely.

CONSTRAINING AGRESSIVE REPORTING

This section focuses on how rules affect a standard's ability to constrain aggressive reporting. By "aggressive" I mean reporting that is biased to produce an outcome consistent with management's incentives. Aggressive reporting could bias elements of the financial statements either upward or downward, need not be "income increasing," and could refer to balance-sheet elements rather than income-statement elements. I focus on the effects of precision on constraint, beginning with research evidence about behavior associated with precise standards, and then discussing research evidence related to imprecise standards.

Constraining Behavior with Precise Standards

Little experimental research examines reporting choices under precise standards, but auditor surveys provide some relevant evidence. The cell labeled "D" in Table 1 shows NET's evidence that auditors are not likely to require their clients to adjust aggressive reporting that has been specifically structured to meet precise standards, because the client can demonstrate compliance with GAAP. These data highlight that precise standards can create targets that managers use to achieve particular accounting objectives. Modifying transactions can be costly, but in some circumstances managers appear to believe that the costs are justified. Auditors responding to NET's survey usually were reluctant to argue "substance over form" when a client clearly complied with precise accounting criteria, even when those criteria were accompanied by qualifiers indicating the criteria do not apply in all circumstances.

As mentioned previously under "communication," one interpretation of the result in cell "C" of Table 1 is that there are situations in which auditors know more about precise accounting rules than do their clients, and therefore are better able to identify the specific rules and precedents that prohibit a client-preferred accounting treatment. However, another explanation is that auditors' negotiating positions are particularly strong when they can point to precise rules that preclude the client's preferred accounting treatment. Consistent with this interpretation, Gibbins et al. (2001) note in their survey of 93 audit partners that unambiguous standards increase auditors' power in auditor/client negotiations. Respondents identified the presence of relevant accounting standards and the audit firms' accounting expertise as key to successfully resolving negotiations.

Overall, these survey studies suggest that a precise standard enhances the negotiating position of either the auditor or the client, but whose position is strengthened depends on whether the transaction is structured. These results imply that, when standard setters decide whether to increase the precision of a standard, they should weigh the benefits of precision that result from auditors rejecting more unstructured aggressive reporting against the costs of precision that result from companies structuring aggressive transactions that auditors accept.

Prior experimental research suggests that precise standards are less effective in constraining aggressive reporting when managers have latitude in interpreting the evidence related to the standard. For example, Cuccia et al. (1995) find that professional tax preparers respond to a more stringent tax practice standard by interpreting evidence more liberally, such that decisions made under a more stringent standard are as aggressive as decisions made under a less stringent standard (see also Johnson 1993). In fact, the justification process itself might offer sufficient latitude, holding constant precision of standard and evidence. Kennedy et al. (1997) provide evidence that auditors view their reporting choices as more justifiable even if they only consult another partner, regardless of the aggressiveness of the decisions or the extent to which they followed the partner's advice. Thus, precise standards are more likely to help auditors reject unstructured aggressive reporting when there is insufficient other l atitude for the client to justify their preferred position.

Constraining Behavior with Imprecise Standards

The top row of Table 1 reports NET's survey evidence about aggressive reporting attempted under imprecise standards. Cell "B" indicates that NET's respondents identified relatively few instances of structured aggressive reporting with respect to imprecise standards. NET believe this finding is driven by the benefits of transaction structuring being less certain to exceed the costs of structuring when the standard is not precise enough to insure auditor approval of the structured transaction. Cell "A" indicates that NET's respondents identified many instances of unstructured aggressive reporting with respect to imprecise standards, and that auditors in those circumstances waived adjustment of the transaction in 61 percent of the cases. Auditors often justified waiving these adjustments because: (1) the transaction was subjective and the auditor could not prove the client's position was incorrect, or (2) because the transaction decreased current-year income and could be viewed as immaterial or conservative with respect to the current year.

Consistent with these survey results, a relatively large experimental literature provides evidence that the aggressiveness of reporting decisions increases with the imprecision of the relevant reporting standard. This basic result has been produced under a variety of experimental approaches, most of which focus on auditors' role in constraining the aggressiveness of their clients' reporting.

Some studies report experiments in which experienced auditors determine appropriate accounting under standards that differ in their precision. For example, Trompeter (1994) varied the precision of accounting standards and evidence by presenting audit partners with three cases: a marketable-security valuation case in which the client-preferred alternative was clearly precluded, and two contingency cases in which SFAS No. 5 indicated the client's position was incorrect, but arguable. Partners were more likely to allow a client's income-increasing accounting treatment in the cases governed by the less precise standard. Similarly, Hronsky and Houghton (2001) provide evidence that Australian auditors with an average of five years of experience were less likely to allow extraordinary treatment under a new standard that required explicitly that extraordinary items be of a nonrecurring nature.

Other studies manipulate in other ways the latitude available in the comparison between standards and evidence. For example, Libby and Kinney (2000) examine audit managers' decisions about quantitatively immaterial proposed adjustments by varying whether the misstatement is subjective or objective and whether it is material qualitatively. Auditors did not require adjustment of either subjective or objective amounts if the adjustment. produced a change that was qualitatively material to their client. Salterio and Koonce (1997) examine audit managers' and partners' decisions about proposed adjustments and vary the unanimity of relevant precedents. Auditors tended to require their clients to make the reporting choice indicated by precedents when the precedents unanimously favored one choice, but tended to allow their clients to make more aggressive reporting choices when precedents were mixed. A tax study by Marchant et al. (1993) considers analogical reasoning used by students and experienced tax preparers when determining whether a court would allow a questionable tax deduction. More experienced tax preparers were more likely to select a precedent based on a poor analogy to their client's situation when the analogy supported the client's position. Ng and Tan (2002) vary whether audit managers are told that authoritative guidance exists with respect to a hypothetical revenue recognition problem. Auditors were more likely to allow the client's aggressive accounting when authoritative guidance was missing and the client's audit committee was weak.

Knapp (1987) performed an experiment that bears directly on whether audit committee members would support auditors in a dispute with client management. Audit committee members were less likely to support the auditors' position in a conflict with management when the relevant accounting rules were judgmental (operationalized as a dispute over the materiality of an item) than when the relevant accounting standard was objective (operationalized as mandatory disclosure of a subsequent event).

Several studies examine the effects of latitude in experimental markets in which students play the role of auditors. These studies differ in how they operationalize latitude. The experiments reported by Schatzberg et al. (1996) and Calegari et at (1998) allow markets to vary according to the extent to which participants are told that auditors agree on the appropriate accounting, while the experiments reported by Mayhew et al. (2001) vary the accuracy of the evidence provided to participants. Results in these studies suggest that auditors are more likely to misreport in the direction favored by their client when the appropriate accounting is uncertain.

Overall, these various experimental approaches provide evidence that the aggressiveness of reporting decisions increases with the imprecision of the relevant reporting standards. However, it is premature to conclude from these studies that reporting behavior is always more aggressive under imprecise standards, because these studies focused on behavior in settings where incentives tended to favor aggressive reporting.

Other experiments vary incentives and examine reporting behavior under imprecise standards. (5) For example, Farmer et al. (1987) find that experienced auditors' reactions to a client's novel accounting approach were influenced by factors like potential for client loss and potential for litigation. Hackenbrack and Nelson (1996) provide evidence that audit seniors' interpretations of the SFAS No. 5 "reasonable estimate" requirement are more conservative for clients that expose the auditor to higher risk of litigation and reputation loss. These studies indicate that incentive-consistent interpretation of imprecise standards does not imply that managers always prefer aggressive disclosure, or that auditors always agree with that preference. Rather, the balance of incentives determines the directional effect on judgment (Lewis 1980).

Balancing Incentives under Imprecise Standards

Changing the balance of incentives is likely to have a greater effect under imprecise standards. Under precise standards, knowledgeable preparers may achieve a high probability of having their auditors approve aggressive reporting by carefully structuring transactions, regardless of general disincentives provided by regulators or courts. However, when the relevant standard is imprecise, the preparer must cope with more uncertainty and ambiguity about outcomes. If the penalties for aggressive reporting are sufficiently severe, preparers facing imprecise standards presumably will report more conservatively to avoid risking those penalties. Consistent with this perspective, Nelson and Kinney (1997) provide experimental evidence that auditors respond to increased ambiguity about the probabilities relevant to a reporting decision by reporting more conservatively than they do when the probabilities are stated precisely. Zimbelman and Waller (1999) report similar findings with students role-playing auditors in exper imental games.

It is important to take .a broad view of the features of the accounting setting that might tilt the balance of incentives toward or away from aggressive reporting. The research literature has focused for some time on auditor incentives based on concerns about client retention, litigation exposure, and potential consulting revenues, as in Carmichael and Swieringa (1968). However, auditors can also face more subtle incentives. For example, Bazerman et al. (1997) suggest that identification with their client's situation might encourage auditors to allow aggressive reporting, while King (2002) provides evidence that identification with an auditor peer group might discourage aggressive reporting.

The balance of incentives can affect behavior unintentionally. For example, Wilks (2002) and Beeler and Hunton (2002) examine "pre-decisional distortion" of audit evidence, in which incentives affect how evidence is viewed, and therefore affect decisions in unintended, unconscious ways. Both studies examine judgments of client financial-health indicators for purposes of making a "going concern judgment." Wilks' (2002) results indicate that audit managers are more likely to make going-concern judgments that agree with a partner's preliminary views, particularly when they are provided with those views prior to evaluating evidence. Beeler and Hunton's (2002) results indicate that audit partners are more likely to rate evidence as supporting continuance of a going concern when they need to retain their client because they "low-balled" the initial audit fee or see high potential for future nonaudit services. (6) Similarly, when auditors anticipate that they will have to justify decisions to someone whose preferenc es are known, experimental evidence suggests that auditors facilitate eventual approval by biasing their search for information (Turner 2001), their weighting of evidence (Peecher 1996), and their audit opinions (Buchman et al. 1996) to be consistent with those preferences.

Finally, it is important to recognize that behavior can also be influenced by incentives that encourage accurate judgment. When justifying the accuracy of decisions, auditors document more relevant information (Koonce et al. 1995), avoid over-weighting recent information (Kennedy 1993), identify more important information (Tan 1995), and perform higher-quality ratio analysis (Ashton 1990; Tan and Kao 1999). Evidence that accuracy-oriented justification requirements affect behavior suggests that incentives need not favor aggressive or conservative reporting, but could be set to favor accuracy.

Constraint: Summary

The research literature indicates that, regardless of the precision of standards, practitioners consciously or unconsciously make financial reports that are consistent with their incentives. Precise standards appear to help auditors discourage aggressive reporting when opportunities for transaction structuring are not available or clients are unaware of precise rules. However, incentive-consistent reporting choices often can be justified with respect to precise standards via transaction structuring or by aggressive interpretation of the evidence that is evaluated and compared to standards' requirements. And, if standards are imprecise, incentive-consistent reporting choices can be justified via aggressive interpretation of standards. Thus, incentive effects should be viewed as pervasive. If standard setters or regulators desire accurate or conservative reporting, they are most likely to achieve it by combining: (1) standards that are imprecise enough to avoid precise safe harbors, thereby allowing incentive-c onsistent interpretation to take place; and (2) vigorous enforcement activity that tilts the balance of incentives away from aggressive reporting and toward accurate or conservative reporting.

CONCLUSION

The research literature suggests several broad conclusions about the incremental effects of additional rules on a standard's ability to communicate clearly and constrain aggressive reporting.

First, a key to accurate communication is striking the right balance between providing enough rules to communicate clearly and not so many rules that practitioners are overwhelmed. Increasing the extent to which the various rules in a standard are related to each other should help avoid overwhelming practitioners with complexity.

Second, a key to constraining aggressive reporting is the role of incentives in determining behavior when latitude exists. When practitioners must choose between alternative accounting or disclosure treatments, precise standards can offer safe harbors via transaction structuring, and therefore may actually reduce regulators' ability to constrain aggressive reporting. Instead, aggressive reporting can be constrained by a combination of incentives for accurate or conservative reporting and standards that are imprecise enough to offer no safe harbors.

Third, communication and constraint may operate at cross purposes under some circumstances, since the detail necessary to communicate accurately can also create opportunities for transaction structuring. In these cases, transaction structuring could be discouraged by basing guidance more on examples than bright lines, and by including "substance over form" provisions that are enforced when transactions are structured in a manner that is inconsistent with economic substance.

Many changes that are currently occurring or contemplated are consistent with the implications of existing research. For example, the FASB is currently considering developing a database to modify the FASB Current Text by referencing and including literature issued by other standard-setting bodies, such as the SEC, EITF, and AcSEC (FASB 2002b). That initiative should decrease the coordinative complexity inherent in existing standards. The FASB is also proposing a more "principles-based" approach to standard setting that would involve fewer rules and exceptions (FASB 2002a). Included in their proposal is a potential "true and fair view" override to require that the accounting for transactions reflect underlying economic substance. These proposals should discourage transaction structuring and provide sufficient latitude for incentives to affect behavior. Finally, the past few years have witnessed a dramatic increase in public concern about aggressive reporting, as well as increased enforcement activity by regula tors. These changes could tilt the balance of practitioners' various incentives toward more accurate or conservative reporting. In combination with a move toward less of a rules-based approach to standard setting, prior research suggests such a shift in incentives would reduce the aggressiveness of financial reporting.

TABLE 1

Percentage of Attempts to Manage Earnings Adjusted by the Auditor, by
Precision of Accounting Standard and Transaction Structuring

                   Transaction Not        Transaction
                     Structured            Structured          Total

Low Precision   Cell A: 115/292 = 39%  Cell B: 7/21 = 33%  122/313 = 39%
High Precision  Cell C: 97/156 = 62%   Cell D: 7/46 = 15%  104/202 = 51%

 Total                  212/448 = 47%         14/67 = 21%  226/515 = 44%

From Nelson et al.'s (2002, Table 5) survey of auditor experiences with
clients who were attempting to manage earnings. Cell formats are: #
attempts adjusted/total # attempts = percent. Precision of relevant
accounting standard was coded as high if the standard used a
quantitative threshold or specifically allowed or disallowed the
accounting treatment. An attemtp was coded as structured if it involved
a change in the timing or nature of a contract, transaction or activity,
as opposed to involving a judgment or estimation process.

(1.) See Imhoff and Thomas (1988) for a notable exception that provides evidence that managers responded to the precise criteria contained in SFAS No. 13 by structuring lease arrangements to qualify for operating lease treatment.

(2.) Caplan and Kirschenheiter (2002), Christensen et al. (2002), and Dye (2002) are recent examples of theoretical studies in accounting that model various effects of the precision of accounting standards. Kaplow (2000) reviews the law literature relevant to: (1) interactions between rule precision, rule complexity, and uncertainty about eventual adjudication; and (2) choosing the timing at which precision is provided.

(3.) I exclude from this review experimental studies that find that differences in accounting method and/or disclosure format affect the judgments of various financial statement users, because such differences could arise from either principles, or rules-based standards. For examples of these studies, see Dyckman (1964) and Hopkins (1996).

(4.) NET coded an accounting criterion relevant to an earnings-management attempt as "precise" if the criterion (1) was quantified, consistent with "bright line" criteria increasing precision, or (2) if the criterion specifically allowed or disallowed the particular accounting treatment, consistent with more rules increasing precision. They coded an earnings-management attempt as structured if it involved a change in the timing or nature of a contract, transaction or activity, as opposed to involving a judgment or estimation process.

(5.) I focus on studies examining how incentives affect behavior, and exclude studies that examine how variations in auditor incentives affect how independent they are perceived to be by various constituencies. An example of the latter studies is Pany and Reckers (1988).

(6.) Phillips (2002) reports pre-decisional distortion of "going-concern" evidence by accounting students. He also provides evidence that the effect of incentives on interpretation of standards happens more after decisions are made, to justify those decisions. This result is of interest because it reduces concern that incentive-driven interpretations affect decisions. However, this result may not generalize to experienced practitioners, who are better able to anticipate that their decisions must be justified with respect to standards.

REFERENCES

Amer, T., K. Hackenbrack, and M. Nelson. 1994. Between-auditor differences in the interpretation of probability phrases. Auditing: A Journal of Practice & Theory 13: 126-136.

___, ___, and ___, 1995. Context-dependence of auditors' interpletations of the SFAS No. 5 probability expressions. Contemporary Accounting Research 12 (1): 25-39.

American Institute of Certified Public Accountants (AICPA). 1992. The Meaning of "Present Fairly in Conformity with Generally Accepted Accounting Principles" in the Independent Auditors' Report. Statement on Auditing Standards No. 69. New York, NY: AICPA.

Asare, S. K, and L. S. McDaniel. 1996. The effects of familiarity with the preparer and task complexity on effectiveness of the audit review process. The Accounting Review 71 (2): 139-159.

Ashton, R. H. 1990. Pressure and performance in accounting decision settings: Paradoxical effects of incentives, feedback and justification. Journal of Accounting Research 28: 148-186.

Bazerman, M. H., K. P. Morgan, and G. F. Loewenstein. 1997. The impossibility of auditor independence. Sloan Management Review (Summer): 89-94.

Beeler, J. D., and J. E. Hunton. 2002. Contingent economic rents: Insidious threats to auditor independence. Advances in Accounting Behavioral Research (forthcoming).

Beresford, D. R. 1999. It's time to simplify accounting standards. Journal of Accountancy 187: 65-67.

Bonner, S. E., and B. L. Lewis. 1990. Determinants of auditor expertise. Journal of Accounting Research: 1-20.

-----. 1994. A model of the effects of audit task complexity. Accounting, Organizations and Society 19 (3): 213-244.

Buchman, T. A., P. E. Tetlock, and R. O. Reed. 1996. Accountability and auditors' judgments about contingent events. Journal of Business Finance & Accounting 23 (3): 379-398.

Calegari, M. J., J. W. Schatzberg, and G. R. Sevcik. 1998. Experimental evidence of differential auditor pricing and reporting strategies. The Accounting Review 73 (2): 254-275.

Caplan, D., and M. Kirschenheiter. 2002. A model of auditing under bright-line accounting standards. Working paper, Iowa State University.

Carmichael, D. R., and R. J. Swieringa. 1968. The compatibility of auditing independence and management services--An identification of issues. The Accounting Review (October): 697-705.

Christensen, P. O., J. S. Demski, and H. Frimor. 2002. Accounting policies in agencies with moral hazard and renegotiation. Journal of Accounting Research 40 (4): 1071-1090.

Cloyd, C. B. 1995. Prior knowledge, information search behaviors, and performance in tax research tasks. The Journal of the American Taxation Association 17: 82-107.

Cuccia, A. D., K. Hackenbrack, and M. W. Nelson. 1995. The ability of professional standards to mitigate aggressive reporting. The Accounting Review 70 (2): 227-248.

Dilla, W. N., and D. N. Stone. 1997. Representations as decision aids: The asymmetric effects of words and numbers on auditors' inherent risk judgments. Decision Sciences 28: 708-743.

Dyckman, T. R. 1964. On the investment decision. The Accounting Review 39 (2): 285-295.

Dye, R. A. 2002. Classifications manipulation and Nash accounting standards. Journal of Accounting Research 40 (4): 1125-1162.

Farmer, T. A., L. E. Rittenberg, and G. M. Trompeter. 1987. An investigation of the impact of economic and organizational factors on auditor independence. Auditing: A Journal of Practice & Theory 7 (1): 1-14.

Financial Accounting Standards Board (FASB). 1975. Accounting for Contingencies. Statement of Financial Accounting Standards No. 5. Norwalk, CT: FASB.

-----. 1976. Accounting for Leases. Statement of Financial Accounting Standards No. 13. Norwalk, CT: FASB.

-----. 1980. Qualitative Characteristics of Accounting Information. Statement of Financial Accounting Concepts No. 2. Norwalk, CT: FASB.

-----. 1992. Accounting for Income Taxes. Statement of Financial Accounting Standards No. 109. Norwalk, CT: FASB.

-----. 1996. Accounting for Transfers and Servicing of Financial Assets and Extinguishments of Liabilities. Statement of Financial Accounting Standards No. 125. Norwalk, CT: FASB.

-----. 2002a. Proposal: Principles-Based Approach to U.S. Standard Setting. File Reference No. 1125-001. Norwalk, CT: FASB.

-----. 2002b. Project Updates: Simplification and Codification Project. Norwalk, CT: FASB.

Gibbins, M. 2001. Incorporating context into the study of judgment and expertise in public accounting. International Journal of Auditing 5 (3).

-----, S. Salterio, and A. Webb. 2001. Evidence about auditor-client management negotiation concerning client's financial reporting. Journal of Accounting Research 39 (3): 534-563.

Hackenbrack, K., and M. W. Nelson. 1996. Auditors' incentives and their application of financial accounting standards. The Accounting Review 71 (1): 43-59.

Hopkins, P. E. 1996. The effect of financial statement classification of hybrid financial instruments on financial analysts' stock price judgments. Journal of Accounting Research 34 (Supplement): 33-50.

Hronsky, J. J. F., and K. A. Houghton. 2001. The meaning of a defined accounting concept: Regulatory changes and the effect on auditor decision making. Accounting, Organizations and Society 26 (2): 123-139.

Imhoff, E. A., and J. K. Thomas. 1988. Economic consequences of accounting standards: The lease disclosure rule change. Journal of Accounting and Economics 10 (4): 277-310.

Johnson, L. M. 1993. An empirical investigation of the effects of advocacy on preparers' evaluations of judicial evidence. Journal of the American Taxation Association: 1-22.

Kachelmeier, S. J., and R. R. King. 2002. Using laboratory experiments to evaluate accounting policy issues. Accounting Horizons 16 (3): 218-232.

Kaplow, L. 2000. General characteristics of rules. In Encyclopedia of Law and Economics, Volume V. The Economics of Crime and Litigation, edited by B. Bouckaert, and G. De Geest, 502-528. Cheltenham, U.K.: Edward Elgar.

Kennedy, J. 1993. Debiasing audit judgment with accountability: A framework and experimental results. Journal of Accounting Research 31 (2): 231-245.

D. N. Kleinmuntz, and M. E. Peecher. 1997. Determinants of the justifiability of performance in ill-structured audit tasks. Journal of Accounting Research 35 (Supplement): 105-130.

King, R. R. 2002. An experimental investigation of self-serving biases in an auditing trust game: The effect of group affiliation. The Accounting Review 77 (2): 265-284.

Knapp, M. C. 1987. An empirical study of audit committee support for auditors involved in technical disputes with client management. The Accounting Review 62 (3): 577-588.

Koonce, L., U. Anderson, and G. Marchant. 1995. Justification of decisions in auditing. Journal of Accounting Research 33 (2): 369-384.

Lewis, B. L. 1980. Expert judgment in auditing: An expected utility approach. Journal of Accounting Research 18 (2): 594-602.

Libby, R., and H-T. Tan. 1994. Modeling the determinants of audit expertise. Accounting, Organizations and Society 19 (8): 701-716.

-----, and W. R. Kinney. 2000. Does mandated audit communication reduce opportunistic corrections to manage earnings to forecasts? The Accounting Review 75 (4): 383-404.

-----, R. Bloomfield, and M. Nelson. 2002. Experimental research in financial accounting. Accounting, Organizations and Society 27(8): 775-810.

Marchant, G., J. Robinson, U. Anderson, and M. Schadewald. 1993. The use of analogy in legal argument: Problem similarity, precedent, and expertise. Organizational Behavior and Human Decision Processes 55 (1): 94-119.

Mayhew, B. W., J. W. Schatzberg, and G. R. Sevcik. 2001. The effect of accounting uncertainty and auditor reputation on auditor objectivity. Auditing: A Journal of Practice & Theory 20 (2): 49-70.

Nelson, M. W., and W. R. Kinney. 1997. The effect of ambiguity on auditors' loss contingency reporting judgments. The Accounting Review 72 (2): 257-274.

-----, J. A. Elliott, and R. L. Tarpley. 2002. Evidence from auditors about managers' and auditors' earnings management decisions. The Accounting Review 77 (Supplement): 175-202.

Ng, T. B., and H-T. Tan. 2002. Effects of authoritative guidance availability and audit committee effectiveness on auditors' judgments in auditor-client negotiations of audit adjustments. Working paper, Nanyang Technological University.

Pany, K., and P. M. J. Reckers. 1988. Within- vs. between-subjects experimental designs: A study of demand effects. Auditing: A Journal of Practice & Theory 7(1): 39-53.

Peecher, M. 1996. The contingent nature of justification demands' influence on auditors' decision processes. Journal of Accounting Research 34 (Spring): 125-135.

Phillips, F. 2002. The distortion of criteria after decision making. Organizational Behavior and Human Decision Processes 88: 769-784.

Reimers, J. L., S. Wheeler, and R. Dusenbury. 1993. The effect of response mode on auditors' control risk assessments. Auditing: A Journal of Practice & Theory 12 (2): 62-78.

Salterio, S., and L. Koonce. 1997. The persuasiveness of audit evidence: The case of accounting policy decisions. Accounting, Organizations and Society 22 (6): 573-587.

Schatzberg, J. W., G. R. Sevcik, and B. P. Shapiro. 1996. Exploratory experimental evidence on independence impairment conditions: aggregate and individual results. Behavioral Research in Accounting 8 (Supplement): 173-195.

Shaw, J. 1995. Audit failure and regulatory overload. Accountancy 115 (1219): 82-83.

Spilker, B. C. 1995. The effects of time pressure and knowledge on key word selection behavior in tax research. The Accounting Review 70 (1): 49-70.

Stone, D. N., and W. N. Dilla. 1994. When numbers are better than words: The joint effects of response representation and experience on inherent risk judgments. Auditing 13: 1-19.

Tan, H-T. 1995. Effects of expectations, prior involvement, and review awareness on memory for audit evidence and judgment. Journal of Accounting Research 33 (1): 113-135.

-----, and A. Kao. 1999. Accountability effects on auditors' performance: The influence of knowledge, problem-solving ability, and task complexity. Journal of Accounting Research 37 (1): 209-223.

-----, T. B-P. Ng, and B. W-Y. Mak. 2002. The effects of task complexity on auditors' performance: The impact of accountability and knowledge. Auditing: A Journal of Practice & Theory 21 (2): 81-95.

Trompeter, G. 1994. The effect of partner compensation schemes and generally accepted accounting principles on audit partner judgment. Auditing: A Journal of Practice & Theory 13: 56-71.

Turner, C. W. 2001. Accountability demands and the auditor's evidence search strategy: The influence of reviewer preferences and the nature of the response (belief vs. action). Journal of Accounting Research 39 (3): 683-706.

Wallsten, T. S, D. V. Budescu, and R. Zwick. 1993. Comparing the calibration and coherence of numerical and verbal probability judgments. Management Science 39 (2): 177-190.

Wilks, T. J. 2002. Predecisional distortion of evidence as a consequence of real-time audit review. The Accounting Review 77 (1): 51-71.

Wood, R. E. 1986. Task complexity: Definition of the construct. Organizational Behavior and Human Decision Processes 37 (1): 59-82.

Zimbelman, M. F., and W. S. Waller. 1999. An experimental investigation of auditor-auditee interaction under ambiguity. Journal of Accounting Research 37: 135-156.

I appreciate comments from Rob Bloomfield, Dennis Caplan, Robert Freeman, Bob Libby, Bob Lipe, Paul Munter, Bob Swieringa, and Hun-Tong Tan. I am grateful for financial support provided by Cornell's Johnson Graduate School of Management.

Mark W. Nelson is a Professor at Cornell University.