- What is the article about? 400 words
Using the words that other people who are not in accounting and auditing field can read and understand
**DO NOT COPY THE ‘SUMMARY’ PARAGRAPH IN THE ARTICLE**
- How did the author do the research? 100 words
Auditing: A Journal of Practice & Theory American Accounting Association Vol. 34, No. 4 DOI: 10.2308/ajpt-51064 November 2015 pp. 157–170 The Effects of Independent Expert Recommendations on Juror Judgments of Auditor Negligence Jonathan H. Grenier, D. Jordan Lowe, Andrew Reffett, and Rick C. Warne SUMMARY: Audit firms claim that they are subject to unreasonable litigation risk and that legal reforms are needed. One frequently proposed reform is to utilize independent (i.e., court-appointed) experts to examine case facts and provide recommendations to the court. This study provides theory and evidence to examine the general effects of such recommendations on jurors’ judgments and, also, more specifically, to inform critics’ concerns that jurors will merely ‘‘rubber stamp’’ independent experts’ recommendations. Results of an experiment indicate that independent experts’ recommendations shift jurors’ judgments in the direction of the recommendation, but that such effects depend on jurors’ perceptions of the experts’ credibility. Further, consistent with critics’ concerns, independent experts’ recommendations reduce jurors’ sensitivity to specific case facts in some, but not all, contexts. Specifically, when the experts conclude that the auditors were negligent, jurors’ negligence judgments are insensitive to variation in specific case facts (the auditors’ use versus nonuse of a specialist), but are sensitive to such variation when the experts conclude that the auditors were not negligent. Implications for theory, practice, and regulation are discussed. Keywords: auditor litigation; auditor negligence; expert recommendations; nonmeritorious judgments. Data Availability: Available from the authors upon request. INTRODUCTION Litigation stemming from audit engagements continues to be a major concern for public accounting firms (Whitehouse 2009; Rasso 2014). Such concern is warranted, as Donelson (2013) estimates that the six largest audit firms incurred approximately $10 billion in Jonathan H. Grenier is an Assistant Professor at Miami University, D. Jordan Lowe is a Professor at Arizona State University, Andrew Reffett is an Associate Professor at Miami University, and Rick C. Warne is an Assistant Professor at the University of Cincinnati. We thank Ann Backof, Billy Brewster, Kathryn Kadous, Karen Kitching, Steve Kaplan, Zoe-Vonna Palmrose, Brad Pomeroy, and Matt Stern for constructive feedback. Editor’s note: Accepted by Richard C. Hatfield. Submitted: November 2013 Accepted: February 2015 Published Online: February 2015 157 litigation-related costs (excluding insurance) from 1999 through 2007. Furthermore, annual litigation costs steadily increased over this time period. Accordingly, many firms consistently claim that they are subject to unreasonable litigation risk and that legal reforms are needed (Donelson 2013). A frequent criticism of the current liability system for auditors (and other professionals) is that the triers of fact (e.g., judges and jurors) lack sufficient domain knowledge (Posner 1995; Penrod and Heuer 1997; Blomquist and Posner 2010). In particular, critics argue that jurors do not understand the complexities of a given profession and therefore cannot accurately assess whether defendants met the standard of due professional care (Menon 1995; Palmrose 2006). While jurors typically hear testimony from both plaintiff and defense experts (i.e., adversarial experts), such testimony often fails to enhance, and can further confuse, jurors’ understanding of complex issues, as adversarial experts often present conflicting conclusions (Brekke, Enko, Clavet, and Seelau 1991). Further, testimony from adversarial experts often is viewed skeptically by jurors who perceive adversarial experts as ‘‘hired guns’’ who lack independence (Ivkovic and Hans 2003). As an intuitive solution to the problems of adversarial experts, scholars across many disciplines have proposed utilizing independent (i.e., court-appointed) experts in negligence lawsuits (e.g., Wrightsman, Kassin, and Willis 1987; Menon 1995; Thorpe, Oelhafen, and Arnold 2000).1 In auditing, former SEC Deputy Chief Accountant Palmrose (2006, 130) proposed creating an Auditing Master Office within the PCAOB, staffed with independent audit experts to ‘‘assess auditor compliance with GAAP and with PCAOB auditing standards following allegations of audit failure.’’ The findings from such investigations could then be made available to the courts. Despite the intuitive appeal of utilizing independent experts in cases of alleged professional negligence, such proposals have been met with opposition (Cecil and Willging 1994; Goss, Worthington, Stallard, and Price 2001). Most notably, legal scholars have expressed concern that jurors will merely ‘‘rubber stamp’’ expert recommendations (Fienberg 1989; Brekke et al. 1991; Relkin 1993; Goss et al. 2001). Essentially, critics contend that utilizing independent experts could usurp jurors’ traditional roles in legal proceedings by effectively replacing their judgments with that of the independent expert (Goss et al. 2001). This study’s objectives are to provide theory and evidence to systematically examine how independent experts’ recommendations affect jurors’ judgments and, more specifically, to inform critics’ concerns that jurors will rely too heavily on independent experts’ recommendations. We examine these questions by taking the novel approach of observing the sensitivity of jurors’ auditor negligence judgments to variations in specific case facts (e.g., whether the auditors used a specialist) in the presence versus absence of experts’ recommendations. If, as critics contend, jurors simply defer to independent experts’ recommendations without considering specific facts, then jurors’ negligence judgments in the presence of an independent expert recommendation should be unaffected by variations in specific case facts.2 As such, observing the sensitivity of jurors’ judgments to variations in specific case facts when jurors are, versus are not, provided with experts’ recommendations tests the validity of critics’ concerns. Based on heuristic-systematic processing research (Chen and Chaiken 1999), we predict that independent expert recommendations will, as intended, shift jurors’ negligence judgments in the direction of the experts’ recommendation. However, consistent with critics’ concerns, we predict that such recommendations will reduce jurors’ sensitivity to other important case facts in some, but not all, circumstances. Specifically, due to their strong motivation to blame auditors (cf. Kadous 1 An example of an actual reform is that several states now utilize panels of healthcare providers to assess the merits of medical malpractice lawsuits prior to trial (Rasor 1993; McCaughey 2005). 2 We acknowledge that the independent experts also have access to ‘‘specific case facts’’ and utilize this information when making a recommendation. 158 Grenier, Lowe, Reffett, and Warne Auditing: A Journal of Practice & Theory November 2015 2000), we expect jurors’ judgments to be insensitive to variation in specific case facts when the experts recommend that the auditors were ‘‘negligent,’’ but to be sensitive to such variation when the experts recommend that the auditors were ‘‘not negligent.’’ We test the hypotheses using a 3 3 2 between-participants experiment in which participants assess the negligence of auditors who failed to detect a material financial statement fraud. We manipulate whether participants are provided with a recommendation from an independent panel of audit experts (Control versus Treatment) and, within Treatment, whether the expert panel finds that the audit firm is Negligent or Not Negligent. We also manipulate Specific Case Facts by manipulating the auditors’ use of a specialist at two levels (No Specialist and Specialist). Results of the experiment support our predictions. First, the expert recommendations significantly shifted participants’ negligence judgments in the direction of the recommendat
ion. However, such effects largely depended on participants’ perceptions of the experts’ credibility (i.e., the effect of recommendations increased with jurors’ perceptions of the experts’ credibility). Second, as predicted, participants’ negligence judgments were insensitive to whether the auditors used a specialist when the expert panel concluded that the auditors were negligent, but were sensitive to this variation when the panel concluded that the auditors were not negligent. These findings suggest that when independent experts’ recommendations are consistent with jurors’ motivations (e.g., when the expert recommends that the auditors were negligent), many jurors defer to the expert recommendation with less consideration of specific case facts. Thus, one of the primary objections to utilizing independent experts—that such proposals would usurp the traditional role of jurors—appears warranted in certain contexts. This study informs proposals in auditing (e.g., Palmrose 2006) and elsewhere (e.g., Menon 1995; Thorpe et al. 2000; McCaughey 2005) to utilize independent experts’ judgment in cases of alleged professional negligence. We consider independent experts’ recommendations to be a strong, albeit imperfect cue of auditor negligence. By providing theory and supporting evidence that jurors shift their negligence judgments in the direction of independent experts’ recommendations, this study suggests, in support of the aforementioned proposals, that independent experts’ recommendations likely will improve jurors’ judgments.3 However, critics’ concern is not that jurors will fail to follow experts’ recommendations, but that jurors will blindly follow experts’ recommendations without considering specific case facts. Our study helps to reconcile mixed results in prior research (Brekke et al. 1991; Cecil and Willging 1994; Cooper and Hall 2000) by delineating conditions under which providing jurors with independent expert recommendations will most likely desensitize jurors to specific case facts (i.e., when jurors have strong preexisting motivations to find for the plaintiff or defendant). This research provides an important incremental contribution to accounting and other disciplines where independent expert recommendations currently are being used or considered. Overall, results of this study are sufficiently favorable to support further consideration of proposals to utilize independent experts, but also highlight two issues for future research. First, it is important to identify methods to ensure jurors remain attentive to specific case facts, particularly when experts’ recommendations are consistent with jurors’ motivations. Second, future research should examine how to ensure that independent experts appear to jurors to be unbiased and credible. These findings have implications for standard setters as the PCAOB consistently indicates that it considers ‘‘relevant academic research’’ in the standard-setting process (e.g., PCAOB 2012). 3 While we believe independent experts’ recommendations provide a strong cue of auditor negligence, a pure normative benchmark for jurors’ auditor negligence judgments does not exist in the study’s experimental setting. As such, definitive conclusions as to whether independent experts’ recommendations improve jurors’ judgments are not possible. The Effects of Independent Expert Recommendations on Juror Judgments of Auditor Negligence 159 Auditing: A Journal of Practice & Theory November 2015 BACKGROUND The use of independent experts in cases of alleged professional negligence is relatively uncommon, with a notable exception being medical malpractice, as several states utilize panels of medical experts to assess the merits of cases prior to trial (Goss et al. 2001; Rasor 1993). Despite the intuitive appeal of leveraging independent experts, legal scholars have expressed concern that doing so will supplant lay jurors’ role as the primary triers of fact in legal proceedings. Specifically, there is concern that jurors will simply defer to independent experts’ conclusions without sufficiently considering specific case facts and/or imparting their own thoughts and views into their verdicts. As such, critics argue that utilizing independent experts could negate the influence of community values in deciding the outcomes of negligence lawsuits if jurors simply defer to experts (Goss et al. 2001). There is limited, indirect, mixed empirical evidence that addresses this issue. Brekke et al. (1991) find that jurors have slightly better memory of testimony from adversarial experts (i.e., those hired either by the defense or plaintiff ) than non-adversarial (i.e., independent) experts, and are less responsive to the content of testimony from independent experts compared to adversarial experts. Van Dusen (1962) examines malpractice cases in the U.S. District Court for Eastern Pennsylvania over a four-year period and finds that the jury’s verdict was consistent with the independent expert’s recommendation in only 12 of 19 cases (63 percent). These studies suggest that independent experts’ recommendations improve jurors’ judgments but that jurors do not blindly defer to the experts’ recommendation. A survey by Cecil and Willging (1994) finds that only two verdicts of 58 cases were inconsistent with the independent experts’ conclusions. Furthermore, Cooper and Hall (2000) report that jurors unanimously side with an independent expert who favors the plaintiff. Yet, when the independent expert favors the defendant, jurors tend to find for the defendants who are individuals, but generally find against defendants who are corporations.4 HYPOTHESES DEVELOPMENT Expert Recommendations The heuristic-systematic processing model specifies two modes of cognition that decision makers utilize to make judgments (Chen, Shechter, and Chaiken 1996; Chen and Chaiken 1999). The systematic mode involves a comprehensive consideration of diagnostic information. The heuristic mode, however, involves the application of simple judgment ‘‘rules,’’ referred to as heuristics, which are easily retrieved from memory. Thus, juror judgments based on the heuristic mode of processing often are significantly influenced by easily processed and understood cues. One heuristic that is particularly relevant to the current study is the belief among novices that experts’ statements can be trusted (Chaiken and Maheswaran 1994; Chen and Chaiken 1999). The extent to which decision makers engage in systematic versus heuristic processing depends on various situational and dispositional factors. In particular, heuristic processing tends to dominate more effortful systematic processing when individuals have low motivation and/or low capacity for cognitively intensive processing (Chaiken and Maheswaran 1994; Bonner 2008). This finding is important because prior research suggests that jurors likely possess low capacity for effortful processing in complex cases of alleged professional negligence (Winter and Greene 2007). Specifically, to create accurate story structures necessary for systematic processing, jurors need to integrate three sources of information: (1) specific information communicated at trial, (2) prior domain knowledge, and (3) prior generic information related to the case (Pennington and Hastie 4 It is important to note that jurors in Cooper and Hall (2000) generally sided with the plaintiff in a control condition that did not include testimony from an independent expert. 160 Grenier, Lowe, Reffett, and Warne Auditing: A Journal of Practice & Theory November 2015 1993). Existing domain-specific knowledge is critical because it helps evaluators organize new case information into a coherent story structure (Brown and Solomon 1990, 1991). Since jurors typically lack preexisting domain-specific knowledge, they struggle to integrate relevant case information into a coherent story structure (Reffett, Brewster, and Ballou 2012). Thus, jurors often resort to heuristic (as opposed to systematic) information processing strategies that rely on easily understood cues, such as plainti
ff losses in cases of alleged auditor negligence (Kadous 2000, 2001; Fo¨ rsterling 2001; Backof 2014). We contend that independent experts’ recommendations (i.e., that the auditors either were negligent or not negligent) would provide an easy cue for jurors to process and understand. Furthermore, as noted above, prior research has identified the belief that experts’ statements can be trusted as a common heuristic (Chaiken and Maheswaran 1994; Chen and Chaiken 1999). We therefore expect that in cases of alleged auditor negligence, juror judgments, which often are the product of heuristic processing, will be significantly influenced by expert recommendations. H1a: Jurors’ negligence judgments will be higher when jurors receive a recommendation from independent audit experts indicating that the auditors were negligent compared to when the court does not utilize expert recommendations. H1b: Jurors’ negligence judgments will be lower when jurors receive a recommendation from independent audit experts indicating that the auditors were not negligent compared to when the court does not utilize expert recommendations. Expert Recommendations and Juror Sensitivity to Specific Case Facts In audit litigation cases, there is concern whether jurors can comprehend the evidence and successfully differentiate among levels of audit quality (Latham and Linville 1998; Kadous 2000; Arel, Jennings, Pany, and Reckers 2012). Based on theory from psychology, we predict that expert recommendations will reduce jurors’ sensitivity to specific case facts and thus impair their ability to distinguish between levels of audit quality. There is a vast literature in psychology demonstrating that individuals are ‘‘cognitive misers,’’ preferring to rely on simple and time-efficient strategies when making decisions (Fiske and Taylor 1984). As discussed in H1a and H1b, expert recommendations afford jurors a simple heuristic for determining whether the auditors were negligent. And, due to the perceived knowledge superiority of the experts, jurors are likely to trust their recommendations. However, based on preference-consistent processing research (Ditto and Lopez 1992; Ditto, Scepansky, Munro, Apanovitch, and Lockhart 1998; Ditto, Munro, Apanovitch, Scepansky, and Lockhart 2003), we contend that whether jurors defer to independent experts’ recommendations depends on the type of recommendation (i.e., negligent or not negligent). When individuals encounter evidence consistent with their preferences, they readily accept it and do not examine the evidence’s validity. Conversely, when individuals encounter evidence inconsistent with their preferences, they question the evidence’s validity by actively seeking preference-consistent evidence.5 Prior research clearly indicates that jurors have a strong preference to find auditors negligent. Kadous (2000, 2001) finds that, in order to maintain a belief in a just world (Lerner and Simmons 5 A medical diagnosis example clarifies the distinction. A patient receiving a clean bill of health readily accepts the diagnosis without questioning the validity of the diagnosis. On the other hand, a patient receiving an unfavorable diagnosis will generate reasons why the diagnosis may be incorrect and actively seek out second opinions in an attempt to refute the unfavorable diagnosis. In accounting, Hales (2007) and Thayer (2011) find that nonprofessional investors tend to engage in preference-consistent processing. The Effects of Independent Expert Recommendations on Juror Judgments of Auditor Negligence 161 Auditing: A Journal of Practice & Theory November 2015 1966), jurors have significant motivations to blame the auditors (Walster 1966), which desensitizes them to audit quality cues when there are severe consequences associated with an undetected fraud. Similarly, research on the audit expectations gap indicates that evaluators typically expect absolute, rather than reasonable, assurance and frequently ask, ‘‘where were the auditors?’’ when material misstatements go undetected (Epstein and Geiger 1994; McEnroe and Martens 2001). Likewise, research on outcome bias finds that jurors are often overly harsh on auditors in auditor negligence lawsuits (cf. Lowe and Reckers 2006; Peecher and Piercey 2008). According to preference-consistent processing, this preference to find the auditors negligent will affect how jurors respond to expert recommendations. In particular, due to preference consistency, we expect that jurors will have lower motivation to question the merits of expert recommendations indicating that the auditors were negligent. As such, jurors who receive a ‘‘negligent’’ recommendation from independent experts are unlikely to thoroughly consider specific case facts, such as whether the auditors used a specialist. On the other hand, due to preference inconsistency, jurors should be less inclined to defer to independent experts’ ‘‘not negligent’’ recommendations. In this case, jurors will be inclined to diligently evaluate and consider specific case facts to potentially disconfirm the preference-inconsistent expert recommendation.6 H2: When jurors receive recommendations from independent experts, jurors’ negligence judgments will be less sensitive to variation in specific case facts when the recommendation is that the auditors were negligent, but will remain sensitive to such variation when the recommendation is that the auditors were not negligent. EXPERIMENTAL DESIGN The study employs a 3 3 2 between-participants experimental design in which jury-eligible individuals assess the negligence of an audit firm that failed to detect a material, fraudulent overstatement of inventory. The experiment manipulates whether participants are provided with the recommendation from a court-appointed panel of independent audit experts (Control versus Treatment) and, within Treatment, whether the panel found that the audit firm was negligent or not negligent. The experiment also manipulates specific case facts by manipulating whether the auditors utilized an inventory specialist to help measure a specialized type of inventory as audit standards recommend (see independent variables). Participants We recruited 346 participants using Amazon’s Mechanical Turk, an online marketplace where ‘‘workers’’ complete online tasks for compensation. Only U.S. citizens of at least 18 years of age were eligible to participate. Participants who correctly answered 90 percent of the comprehension check questions earned $1.00. As participants spent an average of 25 minutes on the task, our average hourly wage of $2.40 is well above the average reservation wage of $1.38/hour on Mechanical Turk (Horton and Chilton 2010).7 6 In the spirit of the scientific method and full disclosure, we acknowledge that this prediction was less precise before collecting or analyzing data. We originally expected expert recommendations to desensitize participants to specific case facts regardless of the nature of the recommendation. However, we subsequently discovered strong, preexisting theory to predict a differential effect based on whether the expert recommendation is negligent or not negligent. 7 Experimental data obtained from Mechanical Turk are considered highly reliable and participant demographics are significantly more diverse and representative of the U.S. population and undergraduate students (Paolacci, Chandler, and Ipeirotis 2010; Horton, Rand, and Zeckhauser 2011; Buhrmester, Kwang, and Gosling 2011; Farrell, Grenier, and Leiby 2015). Accounting studies have also used Mechanical Turk to proxy for jurors (Peecher, Reffett, and Zimbelman 2015; Grenier, Pomeroy, and Stern 2015) and nonprofessional investors (Rennekamp 2012). 162 Grenier, Lowe, Reffett, and Warne Auditing: A Journal of Practice & Theory November 2015 Experimental Procedures Participants access the experimental case online through their Mechanical Turk accounts and are randomly assigned to experimental conditions. We use Reffett et al.’s (2012) adaptation of Kadous’s (2000, 2001) Big Time Gravel case. The case begins with descriptions of important
concepts including material misstatements, reasonable assurance, auditor negligence, and due professional care. Participants are asked related comprehension check questions (adapted from Peecher and Piercey 2008) to emulate court efforts to ensure juror understanding of these concepts. Participants then read about a CPA firm’s (Jones and Company’s) audit of a fictional mining company, the details of a material fraudulent overstatement of inventory that the auditors failed to detect, and the transcript from a subsequent auditor negligence lawsuit. The transcript includes opening and closing statements (which, in the treatment conditions, include the independent experts’ recommendations, as described below) made by both the plaintiff and defense attorneys and testimony by expert witnesses for both the plaintiff and defendant.8 Participants then make several decisions regarding the auditor’s liability for plaintiff losses (described below). The experimental case concludes with the measurement of several process-related measures, manipulation check questions, and demographic questions. Independent Variables We manipulate, at three levels, whether a panel of audit experts performed an investigation of the audit in question. We use a panel of experts to maximize manipulation strength, as having several audit experts reach a consensus is likely to be more compelling and less biased than the recommendation of a single audit expert. In the Control condition, we do not mention an expert panel investigation. Within the two Treatment conditions, the case first informs participants that prior to the start of the negligence trial, a panel of independent audit experts (not affiliated with Jones & Company) would review the facts of the case, examine Jones & Company’s audit work papers, and conduct interviews of both Jones & Company auditors and Big Time Gravel employees. The case then manipulates whether the expert panel determined that the auditors were negligent or were not negligent. In the Negligent recommendation (Not Negligent recommendation) conditions, the plaintiff’s (defendant’s) attorney incorporates the panel’s recommendation that the auditors were negligent (not negligent) into his/her opening and closing statements. Further, when the independent expert panel determined that the auditors were negligent (not negligent), the expert witness for the plaintiff (defendant) calls attention to the panel’s recommendation to support her view that the auditors were negligent (not negligent). We manipulate, at two levels, a specific case fact embedded in the case to indicate lower- or higher-quality audit procedures in an identical manner as Kadous (2000) and Reffett et al. (2012). In Specialist, the auditors employ a specialist to help measure the gravel inventory that is stored in large outdoor piles. Yet, subsequent information reveals that the client bribed the specialist, explaining why the auditors failed to detect the fraud. In No Specialist, the auditors do not consult a specialist to assist in measuring inventory. As employing a specialist is recommended when auditors lack the expertise to precisely measure client inventory (Kadous 2000), these levels represent case facts that should impact jurors’ auditor negligence judgments. However, we contend 8 Following Kadous (2000, 2001), the case incorporates a significant amount of mundane realism in the form of a court transcript that specifies the opening/closing statement and testimony and cross-examination of expert witnesses. This realism is included to promote external validity and to minimize participant perceptions that they are missing key case facts. The Effects of Independent Expert Recommendations on Juror Judgments of Auditor Negligence 163 Auditing: A Journal of Practice & Theory November 2015 that auditor adherence to professional standards is sufficiently ambiguous in either case for either negligent or not negligent juror verdicts to be potentially meritorious.9 Dependent Variables Our primary measure of auditor negligence is participants’ negligence verdicts (Negligent or Not Negligent), which is consistent with prior auditor litigation studies (e.g., Kadous 2001; Cornell, Warne, and Eining 2009; Reffett 2010). We also include a continuous measure of negligence (0 ¼ Certainly not negligent; 100 ¼ Certainly negligent) to examine the robustness of the dichotomous measure. In addition, if a participant finds that the auditors were negligent, then they indicate the amount of compensatory damages (0–$10 million) and punitive damages (0–$10 million) they would require the auditor to pay the plaintiff. It is important to note that previous research indicates that jurors consider different factors (e.g., ability to pay) when assessing damages compared to when assessing negligence (Kadous 2000; Lowe, Reckers, and Whitecotton 2002). In addition, for participants who received a recommendation from an expert panel, we measured participants’ perceptions of the credibility of the panel of independent experts, (0–100 scale; 0 ¼ not credible at all, 100 ¼ extremely credible). RESULTS Manipulation Checks For participants in the Treatment conditions, after the closing statements, 90.9 percent of participants correctly identified the expert panel’s recommendation. For Specific Case Facts, 86.4 percent of participants correctly indicated whether the auditors used an independent specialist to help measure the inventory. The high pass rates indicate that the manipulations were successful. Excluding participants who failed at least one manipulation check question does not affect the inferences from the experiment’s results. Overall Results Table 1 presents descriptive statistics (Panel A, displayed graphically in Figure 1) and a general linear model with a logit link (Panel B) for Verdicts. 10 Consistent with H1a and H1b, the general linear model indicates a significant main effect of Recommendation (v2 ¼ 39.94; p , 0.001), indicating that panel recommendations affect participants’ negligence judgments (as evidenced by the separation of the lines in Figure 1). We also observe a significant main effect of Specific Case Facts (v2 ¼ 21.97; p , 0.001) with participants assessing the auditors as less negligent when the auditors used a specialist (i.e., evidenced by the downward-sloping lines in Figure 1). The significant interaction of Recommendation and Specific Case Facts (v2¼11.39; p¼0.003) provides preliminary support for H2, as it appears that the Negligent recommendation, but not the Not Negligent recommendation, makes jurors less sensitive to Specific Case Facts (as evidenced by the ‘‘flatter’’ line in Negligent relative to Control in Figure 1). Next, we formally test our hypotheses. 9 This view is supported by the findings in Reffett et al.’s (2012) auditor evaluator condition as 4 out of 26 auditor evaluators (15 percent) found that the auditors were negligent in Specialist, whereas 12 out of 25 auditor evaluators (48 percent) found that the auditors were negligent in No Specialist. The fact that these percentages are not 0 or 100 percent indicates that practicing auditors disagree on whether the auditors were negligent, and that either expert recommendation is potentially meritorious. 10 Inferences from our results are unchanged using compensatory or total (compensatory þ punitive) damages or the continuous measure of negligence. Therefore, those dependent variables are not further discussed. The ANOVA for punitive damages (not tabulated) does not reveal any significant terms, suggesting that expert recommendations and audit quality influence judgments of compensatory damages, but not punitive damages. 164 Grenier, Lowe, Reffett, and Warne Auditing: A Journal of Practice & Theory November 2015 Tests of Hypotheses: Effect of Expert Recommendations H1a predicts that participants’ negligence judgments will be higher when they are provided with an expert recommendation indicating that the auditors were negligent. Supporting H1a, the rate of negligent verdicts is significantly higher in Negligent (77/118 ¼ 65.3 percent) compa
red to Control (55/115 ¼ 47.8 percent; v2 ¼ 7.20; pone-tailed ¼ 0.004). However, this effect is entirely driven by the Specialist condition. Within Specialist, the rate of negligent verdicts is significantly TABLE 1 Experimental Results Auditor Negligence Verdicts Panel A: Descriptive Statistics (Verdicts) Recommendation Specific Case Facts No Specialist Specialist Total Control Proportion 40/58 15/57 55/115 Percentage 69.0% 26.3% 47.8% A B Negligent Proportion 39/59 38/59 77/118 Percentage 66.1% 64.4% 65.3% C D Not Negligent Proportion 21/55 6/58 27/113 Percentage 38.2% 10.3% 23.9% E F Panel B: General Linear Model (Logit Link, Binomial Distribution) Source df v2 p-valuetwo-tailed Recommendation 1 39.937 , 0.001 Specific Case Facts 2 21.970 , 0.001 Recommendation 3 Specific Case Facts 2 11.386 0.003 Panel C: Planned Comparisons Contrast v2 p-valueone-tailed H1a (Negligent versus Control) 1/2 (C þ D) . 1/2 (A þ B) þ17.5% 7.20 0.004 H1b (Not Negligent versus Control) 1/2 (E þ F) , 1/2 (A þ B) #23.9% 14.18 , 0.001 H2 (Recommendation ¼ Negligent) C # D , A # B #41.0% 9.59 0.001 H2 (Recommendation ¼ Not Negligent) E # F , A # B #14.9% 0.52 0.409 Variable Definitions: Verdicts ¼ participants that indicated that the auditor was negligent; Recommendation ¼ three-level factor manipulating the recommendation of a panel of audit experts; Control ¼ no recommendation provided; Negligent ¼ recommendation is negligent; Not Negligent ¼ recommendation is not negligent; Specific Case Facts ¼ two-level factor manipulating the use of a specialist; No Specialist ¼ audit firm does not employ an independent specialist to measure gravel inventory; and Specialist ¼ audit firm employs an independent specialist to measure gravel inventory. The Effects of Independent Expert Recommendations on Juror Judgments of Auditor Negligence 165 Auditing: A Journal of Practice & Theory November 2015 higher in Negligent (38/59 ¼ 64.4 percent) compared to Control (15/57 ¼ 26.3 percent; v2 ¼ 16.95; p , 0.001). Within No Specialist, the rate of negligent verdicts is lower in Negligent (39/59 ¼ 66.1 percent) than Control (40/58 ¼ 69.0 percent), but the difference is not significant (v2 ¼ 0.11; p ¼ 0.741). H1b predicts that participants’ negligence judgments will be lower when they are provided with recommendations indicating that the auditors were not negligent. Supporting H1b, the rate of negligent verdicts is significantly lower in Not Negligent (27/113 ¼ 23.9 percent) compared to Control (55/115 ¼ 47.8 percent; v2 ¼ 14.18; p , 0.001). This reduction in negligence verdicts is significant at both levels of Specific Case Facts. Within Specialist, the rate of negligent verdicts is significantly lower in Not Negligent (6/58 ¼ 10.3 percent) compared to Control (15/57 ¼ 26.3 percent; v2 ¼ 10.77; p , 0.001). Likewise, within No Specialist, the rate of negligent verdicts is significantly lower in Not Negligent (21/55 ¼ 38.2 percent) compared to Control (40/58 ¼ 69.0 percent; v2 ¼ 4.91; pone-tailed ¼ 0.013). Overall, results indicate that independent expert recommendations significantly influence jurors’ negligence verdicts (the only exception being when a specialist is not used and the experts recommend that the auditors were negligent, i.e., where jurors are already likely to find the auditors negligent). If one assumes that independent experts’ recommendations likely reflect the merits of the case against the auditors, then results for H1a and H1b suggest that providing jurors with FIGURE 1 Graphical Representation of Results Verdicts ¼ participants that indicated that the auditor was negligent; Recommendation ¼ three-level factor manipulating the recommendation of a panel of audit experts; Control ¼ no recommendation provided; Negligent ¼ recommendation is negligent; Not Negligent ¼ recommendation is not negligent; No Specialist ¼ another cue where the audit firm does not employ an independent specialist to measure gravel inventory; and Specialist ¼ another cue where the audit firm employs an independent specialist to measure gravel inventory. 166 Grenier, Lowe, Reffett, and Warne Auditing: A Journal of Practice & Theory November 2015 independent experts’ recommendations generally would result in more meritorious juror judgments of auditor negligence. H2 predicts that independent experts’ recommendations will make jurors insensitive to variation in specific case facts when the experts recommend that the auditors were negligent, but not when the experts indicate that the auditors were not negligent. Supporting H2, the difference in negligent verdicts between No Specialist and Specialist is significantly lower in Negligent (No Specialist ! Specialist ¼ 1.7 percent) relative to Control (No Specialist ! Specialist ¼ 42.7 percent; v2 ¼ 9.59; pone-tailed ¼ 0.001). However, the difference in negligent verdicts between No Specialist and Specialist is not significantly lower in Not Negligent (No Specialist! Specialist¼27.9 percent) relative to Control (No Specialist! Specialist¼42.7 percent; v2¼0.52; pone-tailed¼0.409). Further, the difference in negligent verdicts between No Specialist and Specialist is significantly higher in Not Negligent relative to Negligent (v2 ¼ 6.23; pone-tailed ¼ 0.006). In summary, results indicate that when independent audit experts indicate that the auditors were negligent, jurors are not attuned to a specific case fact that is diagnostic of the quality of the auditors’ procedures (i.e., whether the auditors used a specialist). Supplemental Analysis We also examine how the perceived credibility of the expert panel moderates the effect of the panel’s recommendations on negligence judgments. In the Negligent condition, Credibility is significantly positively associated with Verdicts (q ¼ 0.51; p , 0.001). Thus, independent experts’ ‘‘negligent’’ recommendations were more likely to result in a negligent verdict as participants’ perceptions of the experts’ credibility increased. Similarly, within Not Negligent, Credibility is significantly negatively associated with Verdicts (q ¼ !0.33; p , 0.001). Thus, independent experts’ ‘‘not negligent’’ recommendations were more likely to result in a not negligent verdict as participants’ perceptions of the experts’ credibility increased. These results indicate that the impact of the expert panel’s negligent and not negligent recommendations depends on evaluators’ perceptions of the panel’s credibility. CONCLUSION This study examines proposals in accounting (e.g., Palmrose 2006) and in other disciplines (e.g., Menon 1995; Thorpe et al. 2000; McCaughey 2005) to utilize independent expert recommendations in cases of alleged professional negligence. Results of this study are generally supportive of these proposals, but highlight the need to address two issues. First, we find some support for a previously expressed concern that jurors will merely ‘‘rubber stamp’’ independent experts’ recommendations without sufficient consideration of specific case facts. Specifically, jurors’ judgments were sensitive to variation in specific case facts (i.e., whether the auditors used a specialist) when the experts concluded that the auditors were not negligent, but were not sensitive when the experts concluded that the auditors were negligent. Second, while recommendations significantly affected jurors’ judgments, such effects largely depended on jurors’ perceptions of the experts’ credibility. Thus, if proposals to provide jurors with independent expert recommendations are implemented, then efforts should be made to maximize experts’ credibility, as jurors will discount or perhaps even ignore both negligent and not negligent recommendations they perceive to come from experts who lack credibility. These results have important implications for courts and standard setters, as audit firms frequently claim that they are subject to unreasonable litigation risk and that legal reforms are needed (Donelson 2013). Our research could assist standard setters as they contemplate solutions to auditor liability concerns for several
reasons. First, given audit firms’ consistent claims that they are subject to unreasonable litigation risk (Donelson 2013), this public-policy concern likely will The Effects of Independent Expert Recommendations on Juror Judgments of Auditor Negligence 167 Auditing: A Journal of Practice & Theory November 2015 persist until solutions are implemented, and our results are sufficiently favorable to warrant further research and consideration. Second, standard setters may benefit from forward-looking ideas. For example, the Cohen Commission proposed mandatory auditor rotation in 1978, and 33 years later, the PCAOB issued a related concept release (PCAOB 2011). Discussions regarding expert panels could occur for years to come since similar panels are already used in the medical profession (e.g., McCaughey 2005). The theory-consistent evidence presented in this study informs such discussions. Our study is subject to limitations along with those typically associated with experimental research. First, although we based our experiment on Palmrose’s (2006) description of how courts could implement expert recommendations, there could be alternative courtroom presentations. For example, courts could restrict attorneys and expert witnesses from either knowing and/or commenting on the recommendations and, rather, have the judge or expert(s) present the recommendations to the jurors. Yet, we are not aware of a theory that would predict that presentation method would change how jurors respond to the recommendations. Second, we examine juror judgments, but negligence judgments are made in deliberating juries. Third, we believe that with proper quality control, expert recommendations would typically reflect the merits of the case (see footnote 2) and examine a scenario where we believe that either a negligent or not negligent recommendation could be meritorious. However, we do not examine how jurors respond to blatant non-meritorious expert recommendations or factors that influence the quality of expert recommendations. We encourage future research to examine these issues. REFERENCES Arel, B., M. Jennings, K. Pany, and P. M. J. Reckers. 2012. Auditor liability: A comparison of judge and juror verdicts. Journal of Accounting and Public Policy 31 (5): 516–532. Backof, A. 2014. The Impact of Audit Evidence Documentation on Jurors’ Negligence Verdicts and Damage Awards. Working paper, University of Virginia. Blomquist, R., and R. Posner. 2010. The Quotable Judge Posner: Selections from Twenty-Five Years of Judicial Opinions. Albany, NY: State University of New York Press. Bonner, S. E. 2008. Judgment and Decision Making in Accounting. Upper Saddle River, NJ: PearsonPrentice Hall, Inc. Brekke, N. J., P. J. Enko, G. Clavet, and E. Seelau. 1991. Of juries and court-appointed experts: The impact of non-adversarial versus adversarial expert testimony. Law and Human Behavior 15 (5): 451–475. Brown, C. E., and I. Solomon. 1990. Auditor configural information processing in control risk assessment. Auditing: A Journal of Practice & Theory 9 (Fall): 17–38. Brown, C. E., and I. Solomon. 1991. Configural information processing in auditing: The role of domainspecific knowledge. The Accounting Review 66 (1): 100–119. Buhrmester, M., T. Kwang, and S. D. Gosling. 2011. Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science 6 (1): 3–5. Cecil, J. S., and T. E. Willging. 1994. Accepting Daubert’s invitation: Defining a role for court-appointed experts in assessing scientific validity. Emory Law Journal 43: 995–1070. Chaiken, S., and D. Maheswaran. 1994. Heuristic processing can bias systematic processing: Effects of source credibility, argument ambiguity, and task importance on attitude judgment. Journal of Personality and Social Psychology 66 (3): 460–473. Chen, S., and S. Chaiken. 1999. The heuristic-systematic model in its broader context. In Dual-Process Theories in Social Psychology, 73–96. New York, NY: The Guilford Press. Chen, S., D. Shechter, and S. Chaiken. 1996. Getting at the truth or getting along: Accuracy-versus impression-motivated heuristic and systematic processing. Journal of Personality and Social Psychology 71 (2): 262–275. 168 Grenier, Lowe, Reffett, and Warne Auditing: A Journal of Practice & Theory November 2015
No comments:
Post a Comment