Blay_CAR_2009
Independence Threats, Litigation Risk, and the
Auditor’s Decision Process*
ALLEN D. BLAY, University of California Riverside
Abstract
This study examines the effect of independence threats and litigation risk on auditors’ eval-uation of information and subsequent reporting choices. Using a Web-based experiment, I tracked auditors’ information gathering and evaluation leading to a going-concern reporting decision. Speci?cally, 48 audit managers assessed client survival likelihood, gathered addi-tional information, and suggested audit report choices. I found that auditors facing high independence threats (fear of losing the client) evaluated information as more indicative of a surviving client and were more likely to suggest an unmodi?ed audit report, consistent with client preferences. In contrast, auditors facing high litigation risk evaluated information as more indicative of a failing client and were more likely to suggest a modi?ed audit report. In addition, the association between risk and report choice was fully mediated by ?nal infor-mation evaluation. This suggests that it is unlikely that different reporting choices resulted from a conscious choice bias, but rather that motivated reasoning during evidence evaluation plays a key role in the effect of risk in auditor decision making.
Keywords Auditor independence; Auditor reporting; Decision process; Going concern;
Motivated reasoning
JEL Descriptors D8, M42
Data Availability Data are available from the author upon request.
Les menaces pour l’indépendance, le risque de litiges
et le processus décisionnel du véri?cateur
Condensé
L’étude expérimentale de l’auteur vise à déterminer si les incitatifs économiques associés àdifférents niveaux de menaces pour l’indépendance du véri?cateur et de risque de litiges in?uent sur l’évaluation de l’information par les véri?cateurs et leurs choix de rapports sub-séquents. L’indépendance du véri?cateur est souvent considérée comme l’atout le plus précieux de sa profession (Johnstone, Sutton et War?eld, 2001). Les autorités de réglemen-tation craignent que l’intérêt personnel soit une menace sérieuse pour l’indépendance du *Accepted by Eric Hirst. This paper is based on my dissertation at the University of Florida. I would like to thank the members of my dissertation committee for their guidance: A. Rashad Abdel-khalik (chair), Stephen Asare, Robert Knechel, and Barry Schlenker. I also thank Kim Sawers, Wendy Bailey, Andy Spicer, and participants from the 2001 American Accounting Asso-ciation Annual Meeting.
Contemporary Accounting Research V ol. 22 No. 4 (Winter 2005) pp. 759–89 ? CAAA
760Contemporary Accounting Research
véri?cateur (par ex., SEC, 2002) ; c’est pourquoi d’importantes restrictions ont été établies dans la loi Sarbanes-Oxley de 2002 (Sarbanes-Oxley Act — SOA), dans le but de prévenir les manquements à l’indépendance du véri?cateur. Parmi les recommandations de la SOA ?gurent la rotation des associés en véri?cation, l’application de limites à certains services autres que la véri?cation, et l’augmentation de l’information communiquée aux comités de véri?cation. Bien que la SOA tente d’atténuer le lien économique entre véri?cateur et client, il n’en reste pas moins que les rentes économiques futures dépendent toujours de la décision du client de continuer de traiter avec le véri?cateur. Ces rentes potentielles créent des incitatifs susceptibles de réduire l’indépendance du véri?cateur (DeAngelo, 1981 ; Magee et Tseng, 1990). Parallèlement, le risque de litiges engendre des incitatifs économiques qui poussent le véri?cateur à maintenir son indépendance, compte tenu de l’importance de la sanction économique qui pourrait découler d’un litige si les participants au marché estiment qu’il n’a pas fait rapport de manière ?dèle (Palmrose, 1988 ; Carcello et Palmrose, 1994 ; Blacconiere et DeFond, 1997). Le risque de litige est donc souvent évoqué comme sauvegarde contre les menaces pour l’indépendance (par ex., Johnstone et al., 2001).
Bien que les auteurs de certaines études récentes exposent des faits con?rmant que les menaces pour l’indépendance in?uent sur le rapport du véri?cateur (par ex., Frankel, Johnson et Nelson, 2002 ; Lacker et Richardson, 2003 ; Chang et Hwang, 2003 ; Ferguson, Seow et Young, 2004), de nombreux autres chercheurs n’ont pu relever aucune donnée démontrant l’existence de brèches systématiques à l’indépendance (par ex., DeFond, Raghunandan et Subramanyam, 2002 ; Chung et Kallapur, 2003 ; Ashbaugh, LaFond et Mayhew, 2003 ; Geiger et Rama, 2003). Les études antérieures traitant d’indépendance du véri?cateur étaient centrées exclusivement sur les décisions ?nales. Il est fréquent que les décisions ?nales suscitent l’intérêt, puisque les allégations de brèches à l’indépendance sont associées à un aboutissement (notamment, la gestion du résultat et les rapports du véri?cateur). Toutefois, même lorsque sont décelés des faits révélant des préoccupations relatives à l’indépendance, la recherche basée uniquement sur l’aboutissement ne permet pas d’établir la distinction entre une distorsion liée au choix, les sujets ajustant leurs critères décisionnels ou choi-sissant un aboutissement conforme aux incitatifs, et une distorsion liée au traitement, les sujets modi?ant le processus d’évaluation des faits et parvenant à des jugements différents.
Cette distinction entre traitement et choix est cruciale pour deux raisons. Premièrement, la décision du véri?cateur étant séquentielle (Knechel et Messier, 1990), les menaces poten-tielles pour l’indépendance et les effets contr?lants du risque de litiges peuvent avoir une incidence sur l’évaluation de l’information même si, comme le donnent à penser les travaux précédents sur l’indépendance des véri?cateurs, les décisions ?nales ne sont pas touchées de fa?on uniforme. Les menaces pour l’indépendance et le risque de litiges engendrent des incitatifs économiques susceptibles de motiver les véri?cateurs à poursuivre des objectifs de rapport orientés et, selon la théorie du raisonnement motivé, les véri?cateurs peuvent évaluer l’information dans le sens d’une conclusion souhaitée (Kunda, 1990). Si l’évaluation de l’information est soumise à l’in?uence de menaces pour l’indépendance ou d’un risque de litiges, des changements dans l’environnement de réglementation pourraient accidentelle-ment réduire l’ef?cacité ou l’ef?cience de la véri?cation par suite de la distorsion du traitement et (ou) d’une recherche d’information élargie ou inef?cace. Deuxièmement, les méthodes visant l’élimination de la distorsion liée au choix et au traitement diffèrent (par ex., Fischhoff, 1982 ; Kennedy, 1993), ce qui a d’importantes conséquences pour les autorités CAR V ol. 22 No. 4 (Winter 2005)
Independence Threats and Litigation Risk761 de réglementation et les praticiens. Ainsi, les efforts visant à favoriser la reddition de comptes dans le but d’améliorer le processus de véri?cation (les normes en matière de documenta-tion de la véri?cation du Public Company Accounting Oversight Board [PCAOB, 2004a], par exemple) sont moins susceptibles d’entra?ner une véritable amélioration de la qualité de la véri?cation si les menaces pour l’indépendance et le risque de litiges donnent lieu à une distorsion liée au traitement de l’information, par opposition à une distorsion liée au choix (Kennedy, 1993).
Pour établir la distinction entre traitement et choix, l’auteur étudie le processus par lequel le véri?cateur effectue le choix d’un rapport pour un client hypothétique faisant face à des dif?cultés ?nancières (qui appellent une décision quant à la continuité de l’exploitation) et contr?le ce processus d’évaluation en analysant le choix ?nal. L’auteur choisit la décision relative à la continuité de l’exploitation pour plusieurs raisons. Premièrement, la norme de véri?cation pertinente, SAS 59 (AICPA, 1988), fournit au véri?cateur peu d’indications en ce qui a trait aux circonstances dans lesquelles l’opinion de ce dernier doit être modi?ée, ne proposant que des critères vagues (comme celui du ? doute substantiel ? — substantial doubt). Deuxièmement, dans ce contexte, le véri?cateur fait face à la fois au risque de litiges (Carcello et Palmrose, 1994) et aux menaces pour l’indépendance, comme celle du risque de révocation (Geiger, Raghunandan et Rama, 1998 ; Chow et Rice, 1982). En?n, plusieurs études d’archives sur les menaces pour l’indépendance ont mis l’accent sur l’information relative à la continuité de l’exploitation dans le rapport du véri?cateur comme étant un domaine qui met en jeu l’indépendance des véri?cateurs (DeFond et al., 2002 ; Geiger et Rama, 2003).
L’auteur procède à une expérience 2 × 2 basée sur le Web, dans laquelle sont manipulés les menaces pour l’indépendance (menace de révocation par le client) et le risque de litiges (avis du service de gestion du risque du cabinet). Quarante-huit véri?cateurs d’échelon équivalent à celui de chef de groupe des Quatre Grands cabinets de services professionnels cherchent de l’information dans une banque de données, à partir de mots clés, pour prendre une décision quant à l’information à donner sur la continuité de l’exploitation. Après analyse de chaque élément d’information, le véri?cateur évalue dans quelle mesure les données recueillies appuient l’af?rmation selon laquelle la société est viable et produit une évaluation globale de la probabilité qu’elle le demeure. Les véri?cateurs peuvent mettre ?n à leur recherche et suggérer un choix de rapport à tout moment.
L’auteur constate que les menaces pour l’indépendance poussent les véri?cateurs àévaluer l’information dans le sens de la conclusion que privilégie le client et que le risque de litiges élevé donne lieu à la tendance opposée. Plus précisément, les véri?cateurs dont l’indépendance est gravement menacée jugent la survie de la société cliente comme étant plus probable, tant au départ qu’après avoir procédé à l’évaluation des nouvelles données. Les véri?cateurs qui font face à un risque de litiges élevé jugent la survie de la société cliente comme étant moins probable après avoir procédé à l’évaluation des données. En outre, les véri?cateurs dont l’indépendance est gravement menacée et pour qui le risque de litiges est lui aussi élevé recherchent et évaluent davantage de données que les véri?cateurs qui font face à toute autre combinaison de ces risques. Ces observations contribuent à la recherche en démontrant que les changements apportés à l’environnement de réglementation qui in?uent sur les menaces pour l’indépendance et sur le risque de litiges peuvent modi?er tant l’évaluation des données par le véri?cateur que la quantité de données qu’il examine.
CAR V ol. 22 No. 4 (Winter 2005)
762Contemporary Accounting Research
En outre, bien que les véri?cateurs qui font face à de graves menaces pour l’indépendance soient davantage susceptibles d’exprimer une opinion non modi?ée, la relation entre les menaces pour l’indépendance et le choix de rapport est pleinement conditionnée par l’éva-luation ?nale des données, ce qui indique que rien ne permet de conclure que les véri?cateurs aient modi?é leurs critères décisionnels, mais que les décisions ont été prises au terme d’un processus décisionnel soumis à une distorsion. Les véri?cateurs qui font face à un risque de litiges élevé sont davantage susceptibles d’exprimer une opinion modi?ée, mais cette rela-tion est aussi pleinement conditionnée par l’évaluation ultime des données. Ces observations contribuent à la recherche en démontrant qu’il y a peu de chances que les menaces pour l’indépendance et le risque de litiges se traduisent simplement par une distorsion liée au choix résultant d’un ajustement des critères décisionnels, mais qu’il s’agit d’une question de jugement et de processus décisionnel. Les efforts récents pour améliorer l’indépendance du véri?cateur, comme l’adoption de la norme de véri?cation 3 du PCAOB portant sur la documentation de la véri?cation (PCAOB, 2004a), sont axés sur l’élargissement de la docu-mentation et de la révision. D’autres efforts, comme la publication de règles de déontologie et d’indépendance proposées en matière d’indépendance, de services ?scaux et d’honoraires conditionnels (PCAOB, 2004b), modi?ent le régime incitatif. La meilleure qualité de la documentation et de la révision est beaucoup plus susceptible de porter fruit s’il s’agit d’un simple problème de distorsion liée au choix, où les véri?cateurs ajustent leurs critères décisionnels, font uniquement des choix conformes aux incitatifs ou déploient des efforts insuf?sants (Kennedy, 1993). Les résultats du conditionnement révèlent qu’il est peu probable que ce soit le cas et que l’établissement de normes futures pourrait nécessiter l’utilisation d’autres méthodes d’amélioration du processus de véri?cation, telles que certaines catégories particulières d’aides à la décision. En outre, si des niveaux d’incitatifs différents in?uent sur l’évaluation et la quantité des données recueillies, l’on ne sait pas clairement comment les ajustements proposés au régime incitatif in?ueront sur la décision du véri?cateur.
1.Introduction
This study experimentally investigates whether economic incentives associated with different levels of auditor independence threats and litigation risk in?uence auditors’ evaluation of information and their subsequent reporting choices. Auditor independence is often cited as the most valuable asset possessed by the auditing profession (Johnstone, Sutton, and War?eld 2001). Regulators are concerned that self-interest is a major threat to auditor independence (e.g., Securities and Exchange Commission [SEC] 2002) and, as a result, the Sarbanes-Oxley Act of 2002 (SOA) codi?ed substantial restrictions designed to protect against lack of auditor independ-ence. The SOA’s recommendations included audit partner rotation, limitations on certain nonaudit services, and increased reporting to audit committees. Although the SOA attempts to minimize the economic bond between the auditor and the cli-ent, future economic rents continue to be provided only if the auditor is retained by the client. Such potential rents create incentives that potentially reduce auditor independence (DeAngelo 1981; Magee and Tseng 1990). At the same time, litiga-tion risk provides economic incentives for the auditor to remain independent because costly litigation may ensue if market participants believe auditors have not faithfully reported (Palmrose 1988; Carcello and Palmrose 1994; Blacconiere and CAR V ol. 22 No. 4 (Winter 2005)
Independence Threats and Litigation Risk763 DeFond 1997). Accordingly, litigation risk is often proposed as a safeguard to mit-igate independence threats (e.g. Johnstone et al. 2001).
Although some recent studies ?nd evidence consistent with independence threats affecting auditor reporting (e.g., Frankel, Johnson, and Nelson 2002; Larcker and Richardson 2003; Chang and Hwang 2003; Ferguson, Seow, and Young 2004), many others fail to detect any evidence of systematic independence violations (e.g., DeFond, Raghunandan, and Subramanyam 2002; Chung and Kal-lapur 2003; Ashbaugh, LaFond, and Mayhew 2003; Geiger and Rama 2003). Prior studies related to auditor independence focus solely on ?nal decisions. Focusing on ?nal decisions is common because allegations of independence violations are associated with outcomes (for example, earnings management and audit reports). However, even when evidence of independence concerns is detected, research based solely on ?nal outcomes is unable to distinguish between a choice bias, where individuals either adjust their decision criteria or choose a ?nal outcome in line with the incentives, and a process bias, where individuals alter the evidential evaluation process and reach different judgements.
This distinction between process and choice is critical for two reasons. First, because an auditor’s decision is sequential (Knechel and Messier 1990), potential independence threats and the controlling effects of litigation risk may affect infor-mation evaluation even if, as prior research on auditor independence suggests, the ?nal decision is not consistently affected. Independence threats and litigation risk provide economic incentives that could motivate auditors to pursue directional reporting goals, and motivated reasoning theory predicts that auditors may evaluate information in a manner consistent with supporting a desired conclusion (Kunda 1990). If information evaluation is affected by independence threats or litigation risk, changes in the regulatory environment could inadvertantly reduce audit effect-iveness or ef?ciency through biased processing and/or extended or ineffective information search. Second, the methods of debiasing choice and process are different (e.g., Fischhoff 1982; Kennedy 1993), which has important implications for regulators and practitioners. For example, accountability-inducing efforts to improve the audit process (e.g., Public Companies Accounting Oversight Board [PCAOB], audit documentation standards 2004a) are less likely to be effective in improving audit quality if independence threats and litigation risk result in an information-processing bias as opposed to a choice bias (Kennedy 1993).
To distinguish between process and choice, I investigate the auditor’s process of reaching a report choice for a hypothetical client facing ?nancial distress (hereafter,“the going-concern decision”) and control for this evaluation process in analyzing the ?nal choice. I choose the going-concern setting for several reasons. First, the relevant auditing standard, Statement on Accounting Standards (SAS) No. 59 (American Institute of Certi?ed Public Accountants [AICPA] 1988), gives the auditor limited guidance about when to modify the audit opinion, providing only vague criteria (“substantial doubt”). Second, in this setting the auditor is faced with both litigation risk (Carcello and Palmrose 1994) and threats to independence, such as dismissal risk (Geiger, Raghunandan, and Rama 1998; Chow and Rice 1982). Finally, several archival studies of independence threats have focused on going-
CAR V ol. 22 No. 4 (Winter 2005)
764Contemporary Accounting Research
concern reporting as an area where auditor independence may be affected (DeFond et al. 2002; Geiger and Rama 2003).
I conducted a 2 × 2 Web-based experiment, manipulating independence threats (threat of dismissal by the client) and litigation risk (opinion of the ?rm’s risk-management department). Forty-eight manager-level auditors from Big 4 pro-fessional service ?rms searched for information from a keyword-based data base to make a going-concern reporting decision. Following each piece of information, the auditors evaluated how strongly the information supported the assertion that the client would remain viable and gave an overall assessment of the client’s likelihood of viability. They could terminate their search at any time and suggest a reporting choice.
I ?nd that threats to independence lead auditors to evaluate information in support of a client-preferred conclusion and that high litigation risk results in the opposite tendency. Speci?cally, auditors with high threats to their independence evaluated their clients as more likely to survive, both initially and as they pro-ceeded to evaluate new information. Auditors facing high litigation risk evaluated clients as less likely to survive at the completion of the evidence evaluation process. Furthermore, auditors facing high levels of both independence threats and litiga-tion risk searched for and evaluated more information than auditors facing any other combination of these risks. These ?ndings contribute to the literature by demonstrating that changes in the regulatory environment that affect independence threats and litigation risk could change both the auditors’ evaluation of evidence and the quantity of evidence they examine.
Moreover, although auditors facing high independence threats were more likely to issue an unmodi?ed opinion, the relation between independence threats and report choice was fully mediated by the ?nal evidence evaluation, indicating that there is no evidence that auditors changed their decision criteria, but that the decisions were reached through a biased decision process. Auditors facing high litigation risk were more likely to issue a modi?ed opinion, but this relation was also fully mediated by the ?nal evidence evaluation. These ?ndings contribute to the literature by demonstrating that the effect of independence threats and litigation risk is unlikely to be a pure choice bias resulting from an adjustment in decision criteria, but a judgement and decision-making process problem. Recent efforts to improve the independent audit, such as PCAOB Auditing Standard No. 3, Audit Documentation (PCAOB 2004a), focus on extended documentation and review. Other efforts, such as Proposed Ethics and Independence Rules Concerning Inde-pendence, Tax Services, and Contingent Fees (PCAOB 2004b), adjust the incentive system. Improved documentation and review is much more likely to be effective in a pure choice bias problem where auditors adjust their decision criteria, make choices solely in line with incentives, or fail to put forth suf?cient effort (Kennedy 1993). The mediation results indicate that this is unlikely to be the case and future standard setting may need to focus on other methods of improving the audit process, such as speci?c classes of decision aids.1 Moreover, if different levels of incentives affect the evaluation of and quantity of evidence gathered, it is unclear how proposed adjustments to the incentive system will affect auditor decision. CAR V ol. 22 No. 4 (Winter 2005)
Independence Threats and Litigation Risk765 The paper is organized as follows. Section 2 discusses prior research and develops hypotheses about the effects of independence threats and litigation risk on an auditor’s evaluation of information and reporting choices. Section 3 describes the experiment and section 4 presents the results. Concluding comments, limitations of this study, and suggestions for future research are in section 5.
2.Background and hypothesis development
Independence threats and litigation risk
Independence risk is the risk to users of the ?nancial statements that an auditor’s impartiality may be compromised (Johnstone et al. 2001). Independence risk is perceived to occur when auditors face threats to their independence. For an inde-pendence threat to exist, a perceived or actual incentive for an auditor to favor a client-preferred reporting choice must be present (Johnstone et al. 2001). Because auditors have an economic bond to their clients (DeAngelo 1981), they have an incentive to be sensitive to client preferences. Regulators are concerned that this economic bond may reduce independence, and hence often focus attention on lim-iting the economic bond (e.g., SEC 2000; SOA 2002). Nonetheless, future eco-nomic rents are contingent on retention by the client, and thus an economic bond continues to exist.
Johnstone et al. (2001) argue that costly litigation (e.g., Palmrose 1988) may serve to mitigate the incentives that lead to concerns about independence. Zhang (1999) presents an analytical model of the auditor’s decision to accept a client-preferred alternative as a decreasing function of litigation risk. In addition, an experiment by Chang and Hwang 2003 found that risks related to potential litigation resulted in a reduction in auditors’ willingness to accept aggressive client-preferred alternatives. Litigation risk is related to client accruals (Lys and Watts 1994) and is also correlated with client-speci?c factors, such as total assets (Lys and Watts 1994; Carcello and Palmrose 1994), as well as ?nancial distress and bankruptcy (St. Pierre and Anderson 1984; Stice 1991; Pratt and Stice 1994). Geiger and Raghunandan (2001) provide evidence that auditors were less likely to modify an audit report for going-concern issues subsequent to the Private Securities Litiga-tion Reform Act of 1995. They argue that the reduction in expected litigation costs to auditors accounted for this shift in reporting decisions.
The in?uence of incentives on auditors’ decisions
Costs related to independence threats and litigation risk have the potential to change reporting decisions by affecting the auditor at both the choice stage and the information-processing stage. Decision makers with similar assessments of a situa-tion often choose decision outcomes on the basis of a weighting of the costs and bene?ts associated with potential outcomes (e.g., Larrick, Nisbett, and Morgan 1993). Davis and Ashton (2002) provide some evidence that auditors may adjust choice thresholds based on the expected costs associated with decision outcomes. Thus, auditors with identical evaluation of evidence could potentially make differ-ent reporting choices that are based on the economic costs associated with the potential outcomes.
CAR V ol. 22 No. 4 (Winter 2005)
766Contemporary Accounting Research
However, in addition to affecting the costs associated with ?nal choices, inde-pendence threats and litigation risk may also in?uence the process through which auditors reach a decision if these factors result in auditors preferring one reporting outcome (for example, having directional goals). Prior research shows that the pro-cess of information evaluation is affected by motivated reasoning — that is, the theory that individuals with directional goals will process information congruent with a desired outcome (Kruglanski and Klar 1987; Kunda 1990, 1999).
Directionally motivated goals arise when an individual is dependent on a par-ticular decision outcome (Neuberg and Fiske 1987; Kunda 1999). This outcome dependence leads individuals to favor the preferences of the entity on which they are dependent. Auditors often have direct and indirect incentives to be ?nancially dependent on a particular outcome. For example, future economic rents, or the potential that future economic rents may be removed, provide auditors with direct incentives to be ?nancially dependent on their clients (e.g., Wallman 1996). In addition, indirect incentives for dependence can result from close professional relationships with the client (Johnstone et al. 2001). Similarly, auditors can face potential ?nancial costs from actual or alleged catering to a client’s preferences from litigation or impairment to reputation (Palmrose 1988; Davis and Simon 1992). In this case, auditors may be ?nancially dependent on external market forces. Thus, auditors are often ?nancially dependent on a particular decision out-come, which leads to directional goals (Kunda 1999).
Prior research in accounting suggests that auditors do have directional goals and tend to reach client-preferred decisions (e.g., Hackenbrack and Nelson 1996). In addition, Kadous, Kennedy, and Peecher (2003) found that goal commitment led auditors to reach client-preferred conclusions, and that requiring auditors to focus on the appropriateness of their choice led to even greater goal commitment. The effect of directional goals on auditors’ evidence evaluation
Ideally, auditors should evaluate all relevant information with accurate judgement as the ultimate goal. This evaluation should not be in?uenced by the potential out-come implied by the judgement (e.g., Beckmann and Gollwitzer 1987). However, motivated reasoning theory posits that decision makers with directionally motiv-ated goals evaluate information consistent with a desired conclusion, so long as the conclusion is justi?able (Kunda 1990; Pyszczynski and Greenberg 1987). This is particularly true when a preference exists prior to the decision stage (Russo, Medvec, and Meloy 1996). Boiney, Kennedy, and Nye (1997) ?nd that motivated reasoning not only in?uences evaluation, but also may lead to choice differences when the ulti-mate choice is ambiguous. In addition, distortion of information continues to occur even when it is in the domain of an individual’s profession and when the decision maker is required to justify his or her decision (Russo, Meloy, and Wilks 2000).
Figure 1 presents a model of the auditor’s reporting decision process. Because this paper focuses on the going-concern reporting decision, the ?gure is presented in that context. However, it can be generalized to other auditor decision settings. The auditor enters the decision setting with a set of incentives that in?uence both directional goals and the decision process. The auditor makes an initial assessment CAR V ol. 22 No. 4 (Winter 2005)
Independence Threats and Litigation Risk767 of the situation, searches for and evaluates information, and terminates this process when suf?cient information has been gathered to generate a ?nal assessment. The auditor’s ?nal assessment is then used to make a reporting decision.
Russo et al. (1996) provide evidence that directional goals in?uence an individ-ual’s initial assessment of a situation. Russo et al. found that individuals evaluated a neutral attribute differently depending on their goals. Later research indicated this was also true for less balanced information (Russo, Medvec, and Meloy 1998). Similarly, because of the incentives presented to auditors by threats to independ-ence and litigation risk, auditors’ directional goals are likely to in?uence their initial evaluations.2 Therefore, I hypothesize the following:
H YPOTHESIS 1a.Auditors facing greater threats to independence will assess
initial information as more supportive of a client-preferred reporting
treatment than auditors facing lower threats to independence.
H YPOTHESIS 1b.Auditors facing higher levels of litigation risk will assess
initial information as less supportive of a client-preferred reporting treat-
ment than auditors facing lower levels of litigation risk.
Hypotheses 1a and 1b predict that auditors will evaluate the same initial evi-dence differently depending on the incentives involved. It is possible, however, that the strength of independence threats and litigation risk is not great enough to lead to statistical differences in initial evaluation if the differences are subconscious. Alternatively, it is possible that auditors merely adjust their initial assessments on the basis of the economic costs involved, but the actual evaluation of gathered evi-dence is not affected. It is therefore important to analyze the evaluation of subse-quently gathered information separately from initial evaluations.
Several factors indicate that incentives may result in differential evaluation of gathered information. For example, Russo et al. (1996) provide evidence that indi-viduals continue to evaluate gathered information in the direction of a desired outcome, which is consistent with research showing that directional goals can lead to the overweighting of evidence consistent with desired outcomes (Lundgren and Prislin 1998), the underweighting of inconsistent evidence (Klayman and Ha 1987), and overall skepticism toward evidence inconsistent with directional goals (Ditto and Lopez 1992; Jain 2003). Because of the incentives presented to auditors by independence threats and litigation risk, directional goals will likely in?uence the evaluation of subsequent information. Therefore, I hypothesize the following:
H YPOTHESIS 2a.Auditors facing greater threats to independence will assess
subsequently gathered information as more supportive of a client-preferred
reporting treatment than auditors facing lower threats to independence.
H YPOTHESIS 2b.Auditors facing higher levels of litigation risk will assess
subsequently gathered information as less supportive of a client-preferred
reporting treatment than auditors facing lower litigation risk.
CAR V ol. 22 No. 4 (Winter 2005)
768
Contemporary Accounting Research
CAR V ol. 22 No. 4 (Winter 2005)
F i g u r e 1L i n k s b e t w e e n i n c e n t i v e s , i n f o r m a t i o n e v a l u a t i o n , a n d a u d i t o r r e p o r t i n g d e c i s i o n s i n t h e a u d i t o r ’s g o i n g -c o n c e r n d e c i s i o n
N o t e s :
T h i s ?g u r e p r e s e n t s t h e h y p o t h e s i z e d m o d e l o f t h e e f f e c t o f i n c e n t i v e s o n t h e a u d i t o r ’s e v a l u a t i o n o f i n f o r m a t i o n a n d r e p o r t i n g c h o i c e s .
L i n k A r e p r e s e n t s t h e h y p o t h e s i z e d r e l a t i o n b e t w e e n i n c e n t i v e s a n d i n i t i a l e v a l u a t i o n s (H y p o t h e s i s 1).
L i n k B r e p r e s e n t s t h e h y p o t h e s i z e d r e l a t i o n b e t w e e n i n c e n t i v e s a n d g a t h e r e d i n f o r m a t i o n (H y p o t h e s i s 2).
L i n k C r e p r e s e n t s t h e h y p o t h e s i z e d r e l a t i o n b e t w e e n ?n a l i n f o r m a t i o n e v a l u a t i o n a n d a u d i t o r r e p o r t i n g d e c i s i o n s (H y p o t h e s i s 3, m e d i a t i o n a n a l y s i s ).
L i n k D r e p r e s e n t s t h e h y p o t h e s i z e d r e l a t i o n b e t w e e n i n c e n t i v e s a n d a u d i t o r r e p o r t i n g d e c i s i o n s (H y p o t h e s i s 3).
Independence Threats and Litigation Risk769 As pictured in Figure 1, on the conclusion of information search and evalua-tion, auditors make a ?nal assessment of the information and make a reporting choice. Hypotheses 2a and 2b predict that the in?uence of independence threats and litigation risk on directional goals will lead auditors to evaluate initial informa-tion and subsequently gathered information in favor of their directional goal. Prior research has shown that directional goals lead to individuals reaching their desired conclusions (Kunda 1990; Russo et al. 1996; Boiney et al. 1997). In addition, inde-pendence threats and litigation risk may cause auditors to adjust their decision thresholds, making it more likely that they reach a decision in line with the economic costs of the outcomes (Davis and Ashton 2002). This suggests the following:
H YPOTHESIS 3a.Auditors facing greater threats to independence will be more
likely to reach a client-preferred reporting decision than auditors facing
lower threats to independence.
H YPOTHESIS 3b.Auditors facing higher litigation risk will be less likely to
reach a client-preferred reporting decision than auditors facing lower
litigation risk.
3.Experiment
Experimental design and independent variables
The experiment is a case that requires experienced auditors to search for informa-tion, evaluate a hypothetical client’s subsequent viability, and suggest an audit opinion on a going-concern decision. I manipulated threats to independence and risk of litigation in a 2 × 2 analysis of variance (ANOVA) between participants design. The instrument was hosted on the World Wide Web.
Consistent with prior literature indicating that the ?nancial bond is closely related to auditor independence (DeAngelo 1981), I operationalized threats to independence as a statement made by the client regarding subsequent retention. Speci?cally, the low independence threat treatment stated that “the original partner on the engagement recently met with management and was informally told that they intend to retain your ?rm as auditors in the future”. The high independence threat treatment indicated that “the original partner on the engagement has told you that management once indicated that they would likely hire a new auditor in the event of a modi?ed audit opinion”.
I operationalized litigation risk as the opinion of the audit ?rm’s risk-manage-ment department and the ownership structure of the client. Speci?cally, the case told participants in the low litigation risk treatment that the client was privately owned and that “risk management consultants at your ?rm indicate that it is unlikely that your ?rm would be participant to litigation as a result of this audit”. The case told participants in the high litigation risk treatment that the company was traded on the National Association of Securities Dealers Automated Quotations (NASDAQ), that the company was followed by analysts, and that “risk manage-ment consultants at your ?rm have cautioned you that the client operates in an environment where litigation against auditors is common”.
CAR V ol. 22 No. 4 (Winter 2005)
770Contemporary Accounting Research
The case used ?nancial statement data from an actual manufacturing com-pany. In designing the case, I sought a going-concern situation that would be ambiguous. To that end, I generated a list of candidate ?rms using the criteria listed in Hopwood, McKeown, and Mutchler 1994 for identifying ?nancially distressed ?rms. To narrow the list, I used Mutchler’s 1983 discriminant model, which has been used to predict going-concern modi?cations with high levels of accuracy. Once the model had identi?ed a set of likely ?rms, I considered those near the cut-off range for modi?cations, looking for ?rms that exhibited a substantial presence of both contrary information and mitigating factors, consistent with the audit approach suggested by SAS No. 59 (AICPA 1988). The search yielded a medium-sized technology ?rm. The disciminant model predicted a going-concern opinion, and the ?rm possessed many contrary and mitigating factors. In addition, the auditors issued an unmodi?ed opinion, and the ?rm continued to exist for the subsequent reporting period. The case was therefore ambiguous.
Participants
Participants were 48 audit managers with a mean experience level of 6.40 years (range 4–12).3 The participants were recruited from multiple of?ces of three Big 4 certi?ed public accountant (CPA) ?rms. I used audit managers because they gener-ally provide substantial input to going-concern decisions. Representatives of the ?rms distributed the case to approximately 75 auditors via an email that included a hyperlink to the case website.4 The number of usable responses received was 48 (three auditors began, but did not complete the case), for an approximate response rate of 64 percent.5 Participation was voluntary. All data were collected prior to the Sarbanes-Oxley Act of 2002.
All participants reported being involved in audit engagements where substan-tial doubt of subsequent viability had existed and the number of such engagements did not differ between treatment conditions (mean = 7.5, F (3,44) = 0.06, p < 0.98). The auditors reported that the majority of these audits resulted in modi?ed opin-ions (mean = 82 percent), and all auditors reported at least one modi?ed opinion.6 On the basis of this information, participants in the experiment appeared to be suf-?ciently experienced to complete the task competently.7
Task and dependent variable measurement
The case required participants to search for and evaluate information for a current year’s audit engagement with “Highpoint Computer Corporation”, a hypothetical client. They were told to assume that they were the primary partner on the engage-ment, that the in-charge auditor had informed them of the possibility that the client might be unable to continue as a going concern, and that their task was to recom-mend an audit opinion.
Each participant ?rst answered background questions related to their experi-ences with fraud, litigation, and client loss.8 Next, the case presented participants with client background information that included a description of the business, bal-ance sheet and income statement information, and the experimental manipulations. Manipulation checks prior to the decision task indicated that the treatments had the CAR V ol. 22 No. 4 (Winter 2005)
Independence Threats and Litigation Risk771 desired effect on the auditors’ assessment of dismissal risk (F (3,44) = 275.77, p< 0.001) and litigation risk (F (3,44) = 92.61, p < 0.001).9 There was no indication of an interaction between the treatment conditions. A disadvantage of performing manipulation checks and asking potentially sensitizing questions prior to the decision task is that it may induce demand effects. However, audit managers are experienced professionals who are unlikely to be in?uenced by experimental demands to provide misleading responses (Bamber 1983). In addition, given that independence is the auditing profession’s most valuable asset (Johnstone et al. 2001), it is not likely that auditors would respond to a known experimental demand to forgo independence. This is not true for litigation risk, because it is more likely that auditors would potentially respond to a demand for increased conservatism. However, the signi?cant ?ndings for independence threats provide some indication that sensitization to the task may not have affected the results.
After the manipulation check, the participants read a set of ?nancial statements for the company. On viewing this initial ?nancial information, the auditors were asked, “Based solely on the information presented above, what do you believe is the likelihood that this company will be able to continue to exist for the subsequent year? (0–100%)?” I refer to this measure as the initial survival probability assess-ment and use it as the dependent variable to test Hypothesis 1. After this initial assessment, the participants searched a data base containing 24 information cues. The information available included ?nancial and non?nancial information listed by auditors as being relevant to the going-concern decision (Mutchler, Hopwood, and McKeown 1997; LaSalle and Anandarajan 1996). Table 1 presents a full list of the information cue topics, the keywords for retrieving the cues, the number of partici-pants who viewed each cue, and whether the information was determined to be positive or negative regarding the future viability of the client.10 Figure 2 presents the Web-based search page and an example of an information cue.
The auditors searched for the information cues using keywords or phrases. They were not provided with a list of available information, but only viewed what they sought out as relevant. This allowed for an unconstrained simulation of infor-mation acquisition and processing. Prior research has shown that requiring auditors to process evidence in a pre-established sequence constrains information search behavior and may result in less effective selection of information (Hoffman, Joe, and Moser 2003). The keywords were determined by the author and another experi-enced auditor with over 10 years’ experience using auditing standards and related words. In addition, the keyword search was pilot-tested by two practicing audit partners, each with over 10 years’ experience, who noti?ed the author if they either were unable to ?nd information or found something unexpected. The keywords and cues were adjusted on the basis of their suggestions.
After retrieving each cue, auditors assessed them on an 11-point scale (from ?5“very negative” to 5 “very positive”) in response to the question, “In relationship to the ability of Highpoint to continue as a going concern, how would you rate the evidence you are viewing on this page?” I refer to this measure as the individual cue evaluation. Participants also reassessed the likelihood that the client would be able to continue in existence for the subsequent year and rated their con?dence in
CAR V ol. 22 No. 4 (Winter 2005)
772
Contemporary Accounting Research
CAR V ol. 22 No. 4 (Winter 2005)
T A B L E
1(T h e t a b l e i s c o n t i n u e d o n t h e n e x t p a g e .)
C u e t o p i c
K e y w o r d s
p a r t i c i p a n t s v i e w i n g d i r e c t i o n o f e v i d e n c e
A c t u a l m e a n e v a l u a t i o n
F i n a n c i n g /l i q u i d i t y
F i n a n c i n g ; l i q u i d i t y ; f u n d s ; l o a n s ; l i e n s ; c r e d i t o r s ; b o r r o w i n g s ; d e b t ; c o v e n a n t s ; d e f a u l t s
38
N e g a t i v e
?1.44
S u b s e q u e n t e v e n t s
S u b s e q u e n t e v e n t s ; 2000; n e x t y e a r ; o u t l o o k ; J a n u a r y ; F e b r u a r y ; M a r c h ; ?r s t -q u a r t e r p r o j e c t i o n s ; f o r e c a s t s ; b u d g e t s
31
N e g a t i v e
?2.61
T r a d i n g s e c u r i t i e s T r a d i n g s e c u r i t i e s ; i n v e s t m e n t s
27
N e g a t i v e
?2.30
L o n g -t e r m d e b t L o n g -t e r m d e b t ; l o n g t e r m
22
P o s i t i v e
0.86
A c q u i s i t i o n
A c q u i s i t i o n ; m e r g e r ; a c q u i r e d ; d i v i s i o n ; p u r c h a s e ; a c q u i s i t i o n s ; a q u i r e d
22
P o s i t i v e
1.77
R e c e i v a b l e s R e c e i v a b l e s ; a /r ; c o l l e c t i o n s ; a c c o u n t s
16
P o s i t i v e
?0.87
M a r k e t s /p r o d u c t s P r o d u c t i o n ; m a n u f a c t u r i n g ; i n v e n t o r y ; m a r k e t s ; p r o d u c t s
13
P o s i t i v e
1.85
I n v e n t o r y I n v e n t o r y ; i n v e n t o r i e s ; t u r n o v e r
13
P o s i t i v e
?0.69
M a n a g e m e n t p l a n s
F o r e c a s t s ; b u d g e t s ; i n t e n t i o n s ; i n t e n d s ; a n t i c i p a t e s ; a n t i c i p a t i o n ; p l a n s ; p r o j e c t i o n s
10
P o s i t i v e
1.00
C o m p e t i t i o n C o m p e t i t o r s ; r i v a l s ; i n d u s t r y ; c o m p e t i t i o n
9
N e g a t i v e
?1.11
A c c r u e d e x p e n s e s A c c r u e d ; p a y a b l e s ; e x p e n s e s
8
N e g a t i v e
?0.37
S o l v e n c y /l e v e r a g e S o l v e n c y ; q u i c k ; a c i d ; c u r r e n t ; l e v e r a g e
7
N e g a t i v e
?0.85
R e s t r u c t u r i n g R e s t r u c t u r i n g ; r e o r g a n i z a t i o n
6
P o s i t i v e
1.00
Independence Threats and Litigation Risk
773
CAR V
ol. 22 No. 4 (Winter 2005)
N o t e :
P a r t i c i p a n t s s e a r c h e d b y k e y w o r d a n d w e r e g i v e n a l i s t o f a v a i l a b l e i n f o r m a t i o n r e l a t e d t o t h e i r s e a r c h e s . T h e r e w e r e 24 p o t e n t i a l i n f o r m a t i o n c u e s a v a i l a b l e a n d p a r t i c i p a n t s w e r e a l l o w e d t o c o n t i n u e s e a r c h i n g u n t i l t h e y d e t e r m i n e d t h e y h a d g a t h e r e d e n o u g h i n f o r m a t i o n . T h e i n f o r m a t i o n i s s o r t e d b y n u m b e r o f t i m e s v i e w e d . P a r t i c i p a n t s e v a l u a t e d t h e i n f o r m a t i o n o n a n 11-p o i n t s c a l e , f r o m ?5 (v e r y n e g a t i v e ) t o 5 (v e r y p o s i t i v e ).
C u e t o p i c
K e y w o r d s
p a r t i c i p a n t s v i e w i n g d i r e c t i o n o f e v i d e n c e
A c t u a l m e a n e v a l u a t i o n
E m p l o y e e s E m p l o y e e s ; m a n a g e m e n t ; c e o ; c f o ; c h i e f ; h u m a n
5
P o s i t i v e
1.60
C a s h g e n e r a t i o n C a s h g e n e r a t i n g
4
P o s i t i v e 1.00
O r d e r b a c k l o g O r d e r ; b o o k ; o r d e r s ; b a c k l o g
3
N e g a t i v e
?0.33
I /S a n d c a s h ?o w s I n c o m e ; ?o w s ; c a s h ; n e t
3
N e g a t i v e
0.00
A s s e t s a l e s
S a l e -l e a s e b a c k ; a s s e t ; d i s c o n t i n u e d ; l e a s e ; d i s p o s a l s ; d i s p o s i t i o n s ; f a c t o r y ; f a c t o r i e s
2
P o s i t i v e
1.00
R e s e a r c h a n d d e v e l o p m e n t R &D ; r e s e a r c h ; d e v e l o p m e n t
2
P o s i t i v e
0.00
M a r k e t i n g a n d d i s t r i b u t i o n M a r k e t i n g ; d i s t r i b u t i o n ; a d v e r t i s i n g
1
P o s i t i v e
0.00
F a c i l i t i e s F a c i l i t y ; f a c i l i t i e s ; ?x e d ; e q u i p m e n t ; p r o p e r t y ; b u i l d i n g s
1
N e g a t i v e
?3.00
L e g a l i n f o r m a t i o n L a w s u i t s ; c o u r t ; c a s e s ; l i t i g a t i o n ; l i t i g a t e ; l e g a l ; p r o c e e d i n g s
P o s i t i v e
0.00
S u p p l i e r s S u p p l i e r s
P o s i t i v e
0.00
P r o ?t a b i l i t y R e t u r n ; r o a ; r o e ; p r o ?t a b i l i t y
N e g a t i v e
0.00
774Contemporary Accounting Research
Figure 2Examples from Web-based experiment
Note:
The experiment described in this paper was conducted on the World Wide Web. Participants searched for information using the keyword search engine shown in panel A. A participant inputting keywords related to debt or ?nancing would be presented with the information cue in panel B. There were 24 information cues available.
Panel A: Search screen
Please search for information that you consider relevant to making your decision about which type of audit report to issue. After you input your search terms, you will be given a listing of all available items which meet your search criteria. At no point during this task should you use the “back” button on your browser. You will be automatically returned to your most recent search following your analysis of each piece of information you view. If you wish to review a previously viewed piece of information, please search for it again. When you have gathered enough information to make your decision on audit report type, click on the “Issue Report” button that will be present on the search page following each piece of information you view. Words less than three letters will be ignored. Also, the
Panel B: Financing and liquidity information cue
?During 1999, Highpoint entered into a new agreement providing for a $17.6 million credit facility that matures on September 1, 2002. As of December 31, 1999, outstanding balances under the credit facility were $14.1 million. The loan is payable in 30 monthly installments of approximately $150,000 beginning January 1, 2000 and ending June 1, 2002. The facility may be repaid and reborrowed at any time without penalty. The com-pany has pledged as collateral substantially all of its assets. In the event of a sale or a sale-leaseback of its largest facility, Highpoint would be required to make a prepayment on the credit facility equal to 75% of the net proceeds from the sale. Management has no other borrowing facilities available at the present time.
?Management expects that the acquisition of its largest competitor and its continued inte-gration of the businesses will improve the company’s liquidity through improved operat-ing performance and the planned disposition of its largest facility. Future liquidity is highly dependent on the revenue growth expected during the upcoming period.
?As of December 31, 1999, the company has $4.3 million in cash on hand, a decrease of $2.6 million from the prior year. However, the company holds publicly traded stock with
a market value of $12.1 million as of December 31, 1999. Management intends to sell
some of this stock for liquidity purposes, should that become necessary.
CAR V ol. 22 No. 4 (Winter 2005)
Independence Threats and Litigation Risk775 that assessment.11 The participants could issue the audit report whenever they believed they had suf?cient evidence to support an opinion. Although the partici-pants reassessed the likelihood of the client’s survival after each information cue, in this paper I focus on only two of these measures: the initial estimate (the initial survival probability assessment) and the ?nal one (the ?nal survival probability assessment), which auditors made just before issuing their report recommenda-tions. Consistent with prior literature (e.g., Bamber 1983; Hirst 1994), I use an inferred likelihood ratio based on the initial and ?nal survival probabilities to measure the effect of information gathering and to test Hypothesis 2. Upon issu-ance of the opinion, the auditors listed the factors that most highly in?uenced their decision, and they answered a debrie?ng questionnaire. The network computer tracked the amount of time spent on each item, as well as overall decision time. 4.Results
Survival assessment with identical information set
Figure 1 presents a diagram of the hypothesized decision model applied to the going-concern setting. Hypothesis 1 predicts main effects for threats to independ-ence (Hypothesis 1a) and litigation risk (Hypothesis 1b) on auditors’ assessment of client viability after viewing only the initial ?nancial information and the manipu-lations. At this stage of the experiment, the manipulations are the only difference among participants. Table 2, panel A presents the mean initial survival probability assessment by risk treatment.
Table 2, panel B presents the results of an analysis of variance (ANOV A) used to compare the treatment effects. The main effect of the independence threat treat-ment was signi?cant (F (1,44) = 7.46, p < 0.01). No main effect was detected for the litigation risk condition (F (1,44) = 1.51, p= 0.23), nor was there an interaction (F (1,44) = 0.30, p= 0.59). In addition, post hoc comparisons of differences between cells indicate that auditors in the treatment condition with high independence threat and low litigation risk evaluated the initial survival likelihood (mean = 72.92 percent) to be signi?cantly higher than did auditors in both the low independence threat/high litigation risk condition (mean = 60.82 percent, F (1,44) = 7.85, p < 0.01) and the low independence threat/low litigation risk condition (mean = 62.92 per-cent, F (1,44) = 5.37, p < 0.03). No other treatment cells had signi?cantly different means from each other. Thus, the data support Hypothesis 1a (independence threats), but not Hypothesis 1b (litigation risk).
Evaluation of gathered information
The experimental design enabled participants to search for and view information based on keywords. Hypothesis 2 predicts that auditors will evaluate gathered evi-dence differently across treatment conditions. To test Hypothesis 2, I used the inferred Bayesian log-likelihood ratio of the ?nal survival probability assessment and the initial survival probability assessment as the dependent variable. Likelihood ratios are commonly used to measure information content (e.g., Bamber 1983; Hirst 1994). Differences in likelihood ratios between conditions imply that the net effect of the gathered information was different.12
CAR V ol. 22 No. 4 (Winter 2005)
776Contemporary Accounting Research
CAR V ol. 22 No. 4 (Winter 2005)
TABLE 2
Analysis of initial survival probability assessment by independence threat and
Panel B: ANOV A Independence threat 833.337.4610.01Litigation risk 168.77 1.5110.23Interaction 33.330.30
10.59
Error
4,912.544
2*
Initial survival probability assessment was made on the basis of a set of ?nancial statements presented prior to any review of information cues. The measure was elicited using the question, “Based solely on the information presented above, what do you believe is the likelihood that this company will be able to continue to exist for the subsequent year? (0–100%)?”?
Independence threat treatment conditions:
High: “The original partner on the engagement has told you that management once indicated that they would likely hire a new auditor in the event of a modi?ed audit opinion.”
Low: “The original partner on the engagement recently met with management and was informally told that they intend to retain your ?rm as auditors in the future.”?
Litigation risk treatment conditions:
High: publicly traded, “Risk management consultants at your ?rm have cautioned you that the client operates in an environment where litigation against auditors is common.”
Low: privately held, “Risk management consultants at your ?rm indicate that it is unlikely that your ?rm would be a participant to litigation as a result of this audit.”
Mean (s.d.)n Independence threat ?
High Low Overall L i t i g a t i o n r i s k ?
High
67.50(8.12)1260.82(9.96)1264.17(9.52)24Low
72.92(10.54)1262.92(13.05)1267.92(12.68)24Overall
70.21(9.61)2461.88(11.40)24
66.04(11.25)48
Effect
Sum of squares F df Prob.
Independence Threats and Litigation Risk777 Table 3 presents analyses comparing the likelihood ratio across treatment con-ditions. Panel A presents descriptive statistics regarding the variables in this analysis. Extreme measures of the likelihood ratio indicate greater adjustments to probabil-ity assessments. More positive values indicate greater increases in the assessments of ?rm survival; more negative values indicate greater decreases. Thus, the high independence threat/low litigation risk group increased their assessments of ?rm survival the most in response to their evaluation of gathered evidence (M = 0.15), whereas the low independence threat/high litigation risk group decreased their assessments the most (M=?0.69).
The results of the ANOV A (Table 3, panel B) indicate a main effect for both independence threat (F (1,44) = 3.57, p = 0.06) and litigation risk (F (1,44) = 7.53, p = 0.01).The greater the independence threat, the less auditors decreased (on average) their predictions that the ?rm would survive; conversely, the greater the litigation risk, the more they decreased ?nal survival probability assessments. These results support both Hypotheses 2a and 2b. Because participants’ informa-tion sets varied for all tests related to gathered evidence evaluation and report choice, I conducted a sensitivity check to control for the quantity of information and the percentage of positive information viewed by the auditors by including these data as covariates in the analysis. Although percentage of positive informa-tion was signi?cantly related to the likelihood ratios (p< 0.01), the quantity of information viewed was not statistically related (p< 0.75). Independence threats and litigation risk continued to be signi?cantly related to the likelihood ratios. Thus, including controls for the quantity and overall direction of the information gathered did not affect the results presented here.
Additional analyses
To provide additional support for my assertion that the directional goals implied by the treatment conditions were driving differences in the evaluation of information, I also analyzed several measures of the auditors’ individual cue evaluations.
Auditors evaluated each cue viewed on a scale of ?5 (“very negative”) to 5 (“very positive”). The mean individual cue evaluation was ?0.52, indicating that on average auditors evaluated the additional information as mildly negative. An untabulated analysis of the mean individual cue evaluations by treatment condition indicated that increased threats to independence led auditors to evaluate individual cues as more indicative of subsequent viability (F (1,239) = 8.47, p < 0.01). Con-versely, increased litigation risk led auditors to evaluate individual cues as less indicative of subsequent viability (F (1,239) = 4.58, p = 0.03). This is consistent with the notion that directional goals in?uence the evaluation of individual evi-dence and provides additional support for Hypotheses 2a and 2b.13
Overall assessment of going-concern status and report issuance
To test the overall effect of initial information and information gathering, I used an ANOVA to compare participants’ ?nal survival probability assessments across treatment conditions (Table 4). The ANOV A revealed main effects for both inde-pendence threat (F (1,44) = 9.86, p < 0.01) and litigation risk (F (1,44) = 7.82,
CAR V ol. 22 No. 4 (Winter 2005)
778Contemporary Accounting Research
TABLE 3
Analysis of inferred log-likelihood ratios for evaluation of gathered information by
Extreme measures of the likelihood ratio indicate greater adjustments to probability assessments. More positive values indicate greater increases in the assessments;
more negative values indicate greater decreases.
?See Table 2 for descriptions of treatment conditions.
CAR V ol. 22 No. 4 (Winter 2005)