CPAs' awareness of litigation risk.
This study examines CPAs' awareness of the litigation risk imposed by audit services. Using CPAs from selected states, I compare perceptions of litigation risk to three objective measures of litigation risk. The first measure is an important legal provision of audit litigation which is assumed to influence litigation risk. Litigation activity reported by the Department of Justice are used for the other two objective measures: frequency of litigation and the level of damages.
Based on these comparisons, I find that the CPAs are not well informed about the litigation risk they face for audit services. Although CPAs provided perceptions consistent with the allegedly important legal standard, a strong majority of CPAs had perceptions that were inconsistent with the Department of Justice statistics about litigation frequency and damage awards. In each case, the misperceptions followed a systematic pattern.
This finding suggests that CPAs may be facing more risk than they thought or incurring more costs than necessary. The finding calls into question the effectiveness of the civil liability system to deter audit improprieties and leads to several questions for future research.
Audit litigation is an important topic to practitioners and academicians as evidenced by the number of articles in their literature (see Latham and Linville, 1998 and Palmrose, 1997 for a review of the literature). Reinforcing the idea that audit litigation is important to practicing CPAs, the American Institute of Certified Public Accountants (AICPA) and the state societies of Certified Public Accountants labored to ease litigation risk faced by CPAs (von Brachel, 1996). These efforts included successful lobbying of Congress for the passage of the Private Securities Reform Act of 1995 and the Securities Litigation Uniform Standards Act of 1998. The effectiveness of these reforms to curb lawsuits against the accounting profession is still an unanswered question (see Investor Relations Business of February 1, 1999 and March 29, 1999 for reports of increasing securities litigation) which suggests that litigation remains a topic of interest.
One issue that has not been addressed in the literature is the accuracy of the CPAs' perceptions of the litigation risk imposed by performing audits. Latham and Linville (1998) discuss the purposes of civil liability and one important purpose is deterrence. By imposing a potential cost through civil liability, the CPA should be reluctant to perform an inappropriate act. However, if the CPA has no understanding of the potential liability or a mistaken understanding of the potential risk, much of the desired deterrence effect could be lost or be counter-productive. The purpose of this paper is to determine if CPAs have accurate perceptions of the liability imposed on them by auditing. The accuracy of the CPAs' perceptions is determined by comparing the perceptions to three objective standards.
In comparing perceptions and objective measures of the litigation risk imposed, I find that CPAs have little understanding of the litigation risk present. Although their perceptions about litigation risk correspond to an (allegedly) important legal standard in audit litigation, their perceptions are inconsistent with the measures of litigation severity provided by the Department of Justice. As discussed later, unintended costs could be imposed on the individual CPAs and society by the CPA's misunderstanding of the litigation risk imposed by auditing services.
The remainder of this paper is organized as follows. In the next section, I discuss the collection of the perceptions about litigation risk and issues of the research instrument. In the third section, I discuss the determination of objective standards. Also, in this section, I discuss the process for designating perceptions as accurate or inaccurate. In the fourth section, the results of the study are presented. The paper concludes with a discussion which includes limitations of the study and suggestions for future research.
COLLECTION OF PERCEPTIONS ABOUT LITIGATION RISK
The data used in this study were collected by survey from CPAs in six states for a study on third-party liability standards. Courts use third-party liability standards to determine which third-parties have legal standing to sue a CPA for allegedly negligent audits. Third-party liability standards are considered to be a major determinant on litigation risk (American Institute of Certified Public Accountants, 1995). Data were collected from CPAs in states which provided extremes in third-party standards (the specific standards are discussed later). A standard imposing a lower level of litigation risk was used by three states (New York, Illinois, and Kansas) and standard imposing a higher level of litigation risk was used in three states at the time of data collection (New Jersey, California, and Wisconsin). In order to provide controls for the original study, only CPA firms with five to twelve CPAs were surveyed. The data was collected from the CPAs in early 1992 about their operations and perceptions for 1991.
Eight-hundred and twenty-eight (828) research instruments were distributed. Of these, 207 were returned for an overall response rate of 25%. This response rate is comparable to other accounting studies (Dalton, Hill, and Ramsey, 1994; McEnroe and Martens, 2001). Of the instruments returned, several did not provide responses to all inquiries resulting in 191 to 196 (23.07% to 23.67%) usable responses to the research instrument questions. Of these observations, 11 were from Kansas which had no counties in the Department of Justice studies and thus, are dropped from the study resulting in a final sample size for this study of 180 to 185. A comparison of the latest respondents to the earliest revealed no response bias.
In the research instrument, the CPAs were asked to provide their perceptions of the litigation risk they faced. A seven-point Likert scale was used to capture their responses to a series of five questions which broke litigation risk into separate components. All five questions are listed in Table 1. The responses to these five questions are compared to the litigation risk assume to be imposed by the third-party liability standard used in the state. The responses to the questions about the likelihood of a lawsuit and the potential damages (questions 3 and 4 from Table 1, respectively) were also compared to reported litigation statistics from the Department of Justice as described later.
OBJECTIVE STANDARDS OF LITIGATION RISK
The perceptions of the CPAs are compared to three objective standards. One standard is based on an (allegedly) important legal precedent existing at the time of the data collection and the other two (frequency of lawsuits and damages awarded) are based on the surveys of the U. S. Department of Justice's Bureau of Justice Statistics (Department of Justice, 1995a and 1995b) (Henceforth, DOJ). For convenience, a perception that is consistent with the objective standard is called "accurate" and a perception that is not consistent with the objective standard is "inaccurate."
The first objective standard of litigation risk is the third-party liability standard used in the state which the CPA works. Third-party liability standards are used by the courts to determine which third-parties have the right to sue the CPA for alleged audit malfeasance. The foreseeable standard allows a third-party the right to sue the CPA for unintentional audit deficiencies if that third-party was foreseeable by the CPA. The privity standard allows a third-party the right to sue the CPA for unintentional audit deficiencies if the third-party was a contracting party with the CPA or a user designated by the party contracting with the CPA (technically, a near-privity standard but which for convenience is combined with privity). Clearly, the foreseeable standard allows a larger set of potential litigants than the privity standard and thus is assumed to increase litigation risk (Nelson, Ronen, and White, 1988; Lys and Watts, 1994).
The second objective standard is the frequency of litigation which is determined by DOJ statistics. In the report entitled Tort Cases in Large Counties (Department of Justice, 1995a), the Bureau of Justice Statistics of the Department of Justice reports on the tort cases disposed from July 1, 1991 to June 30, 1992 in 45 randomly-selected counties from among the nation's 75 largest counties. Five of the states from which perceptions were collected have counties that are included in the study: California (Alameda, Contra Costa, Fresno, Los Angeles, Orange, San Bernadino, San Francisco, Santa Clara, and Ventura), Illinois (Cook and DuPage), New Jersey (Bergen, Essex, and Middlesex), New York (New York), and Wisconsin (Milwaukee). The same counties are used in a pair of follow-up studies about litigation in 1996 (Department of Justice, 1999; Department of Justice, 2000).
The report entitled Tort Cases in Large Counties (Department of Justice, 1995a) does not list separately tort actions against CPAs. To isolate the tort actions which would include the actions against CPAs, the reported actions for automobile, premises, product liability, medical malpractice, and toxic waste lawsuits are subtracted from the total tort actions reported. The resulting classification would include audit litigation and other unspecified, though presumably related, litigation. The number of actions is then scaled by population to arrive at an incident rate per 100,000 population. A median for each state is determined and compared to the national median to determine if the state has a higher or lower frequency of litigation than the country as a whole.
The third objective standard is the level of damages awarded. In the report entitled Civil Jury Cases and Verdicts in Large Counties (Department of Justice, 1995b), the median amount of final award amounts for civil jury trial cases are reported for the same 45 counties for which litigation frequency data were collected. A median is determined for each state and compared to the national median to determine if the state has a higher or lower median damages than the country as a whole.
I use the following procedure to compare the CPAs' perceptions about the litigation risk to the DOJ data. The perceptions were recorded on a 7-point Likert scale with 4 as the midpoint (see Table 1). All the questions ask for a relative evaluation. Using the national median as the norm which is represented as the midpoint on the Likert scale (= 4), accurate and inaccurate perceptions are determined by approximately where the state median would fall on the Likert scale.
For example, the national frequency of litigation per 100,000 population is 49 and California's frequency is 53. Since California's rate is close to but slightly above the national rate, either a perception of 4 or 5 will be deemed a accurate perception. In all cases except one, a accurate response will occupy two spaces on the Likert scale. In the one exceptional case, the median damage award of New York state is so much greater than the national median that using only the highest point on the scale seemed appropriate.
Each test compares the perceptions of the CPAs about their litigation risk to objective measures of the litigation risk in their states. In the first test which compares the perceptions to the third-party liability standard, I use a simple t-test of means to determine if perceptions differ. The final two tests compare perceptions to the DOJ data. I use a binomial test for these two comparisons because each perception is coded as either accurate or inaccurate. When the number of accurate perceptions is not significantly different than the number of inaccurate perceptions (approximately a 50-50 proposition), I conclude that the CPAs are collectively ignorant of the specific litigation risk. When the number of accurate perceptions is significantly higher (lower) than the number of inaccurate perceptions, I conclude that the CPAs have the correct (incorrect) assessment of the specific litigation risk. All reported p-values are one-tail.
The more inclusive third-party standards (i.e. foreseeable standard) are asserted to impose more litigation risk on CPAs (AICPA, 1995). The perceptions of the CPAs are consistent with this assertion. In response to each of the five questions asked and as predicted, the subjects in the foreseeable states perceived significantly (p = 0.0156 or better, one tail, t-test of means) more litigation risk than the subjects in the privity states.
After the data were collected, California and New Jersey abandoned the foreseeable standard. California's change was accomplished in 1992 with the Bily v. Authur Young & Co. (Bily, 1992) decision of the California supreme court and New Jersey's by legislation in 1995 (Demery, 1996). Consistent with idea that greater inclusiveness of the third-party standards leads to a higher litigation risk, these moves should result in lower litigation rates. To test this idea, I compared the litigation frequency in the California and New Jersey counties listed in the DOJ studies before and after the change to privity. Since the 1992 study reports lawsuit filings and the 1996 study reports court resolutions, the 1992 data are converted to approximate resolutions by using the reported 3% rate of trial resolutions per filing (Department of Justice, 1995b, page 2). A comparison of the litigation frequency between the two time periods is shown in Table 4. Although the litigation frequency is lower in the total sample as predicted, the difference does not achieve statistical significance (p = 0.2295, one tail, t-test of means).
Although this analysis does not support the theory that third-party liability standards are important determinants of audit litigation frequency, a confounding factor may cloud this analysis. In 1995, both houses of the U. S. Congress voted to override President Clinton's veto and allow final enactment of the Private Securities Litigation Reform Act of 1995 (Black, Weissman, and Kroll, 1996). This act was designed to reduce the "abusive" lawsuits filed under the Securities Act of 1933 and the Securities Exchange Act of 1934 (Black, Weissman, and Kroll, 1996). By making actions under federal securities laws less attractive, plaintiffs began to file lawsuits in state court using state securities laws (which were often modeled on the original federal statutes and not yet amended to reflect the 1995 changes) or tort provisions (Kou, 1998). These actions ultimately led to the passage of the Securities Litigation Uniform Standards Act of 1998 which limited the use of state courts to try securities lawsuits but for the three-year period between the passage of these two acts, the state courts remained a viable option for the plaintiff wishing to file a securities lawsuit (Casey, 1999). Therefore, the 1996 frequency rate may be inflated by this rush to the state court houses.
This rush to the state courts may be particularly influential in the New Jersey counties. Since court cases take time to resolve, many of the cases listed in the 1996 data may have been filed under the old legal standard (foreseeability). Plaintiffs and their attorneys would have had two legislative deliberations to monitor in early 1995: in the New Jersey legislature and in the U. S. Congress. Both would have suggested a quick filing in the New Jersey state court system as a prudent action to protect the plaintiff's legal claims.
The lower resolution rate per capita in California in 1996 as compared to the 1992 rate approaches traditional statistical significance (p = 0.1299, one tail, t-test of means). If the 1996 rate is inflated, even slightly, by this rush to the state courts, statistical significance could be lost when an uninflated rate would have allowed the difference to achieve statistical significance.
The second objective standard used is the frequency of litigation. Using data from Tort Cases in Large Counties (Department of Justice, 1995a), the frequency of lawsuits per 100,000 population is calculated. A national median is calculated (49 lawsuits per 100,000 population) and compared to the states' calculated rates to determine if the state has a higher or lower litigation rate than the national rate. Accurate perceptions are compared to inaccurate perceptions to determine if the CPAs' perceptions about litigation frequency are accurate. The results of this test are reported in Table 5.
In the total sample, 53 CPAs had perceptions about litigation frequency consistent with the DOJ data and 131 CPAs had perceptions different than the DOJ data. Overall, the CPAs in this sample are significantly inaccurate (p < 0.0001) about the litigation frequency in their states. Further analysis reinforces this finding. In only Wisconsin do a majority of the CPAs have perceptions consistent with the DOJ findings but this majority does not achieve statistical significance. In California, Illinois, and New York, the majority of the CPAs have perceptions that are inconsistent with the DOJ data (p = 0.0029, 0.0031, and 0.000, respectively). In other words, the CPAs in these states have the inaccurate assessment of the litigation frequency in their states.
Among the CPAs that are inaccurate about their states' litigation frequency, all 131 CPAs overestimated their states' litigation frequency. This consistent pattern of overestimation of the litigation frequency is discussed later and may influence the CPAs' performance on their audits.
The third objective standard used for comparison is damages awards. Overall, 76 CPAs had perceptions about the damages awarded in their states consistent with the DOJ data while 108 CPAs had perceptions that were inconsistent with the DOJ data (p = 0.0036). The CPAs in California and Illinois had perceptions about damages that collectively agreed with the DOJ data (p = 0.0088 and 0.0031, respectively) while the CPAs, as a group, in each of the other states were significantly inaccurate in their perceptions of damages: New Jersey (p < 0.0001), Wisconsin (p = 0.0029), and New York (p < 0.0001).
Analysis of the inaccurate perceptions reveals some stark misconceptions. In all the states, a consistent pattern of inaccurate perceptions develops which each achieve significance. Three of the states were too high in their perceptions and two were too low. All the CPAs in New Jersey (p < 0.0001) and Wisconsin (p = 0.0005) reported perceptions that were too high while all the CPAs in New York reported perceptions that were too low. The perceptions of the CPAs in California and Illinois were mixed but the CPAs in California collectively had perceptions that are too low (p = 0.0373) and the CPAs in Illinois collectively had perceptions that are too high (p = 0.0313).
Audit litigation is an important topic today but probably even more important in the early 1990s prior to substantial legal reforms adopted in the mid-1990s. The data of this study come from the early 1990s. Prior research assumed that CPAs were motivated to reduce litigation risk to a reasonable level. Yet, for avoidance of risk to occur, the CPAs who might take such actions must be aware of that risk and for proper actions to occur, the CPAs must have accurate assessments of the risk. This study compares the perceptions of CPAs to objective standards of litigation risk.
Overall, the CPAs in this study did not seem knowledgeable about the litigation risk that they faced. In one broad measure of potential litigation risk, third-party liability standards, CPAs in foreseeable states recognized that they faced more litigation risk than the CPAs in a privity state. Although this measure has been used to determine litigation risk in several studies (Nelson, Ronen, and White, 1988; Lys and Watts, 1994), other legal provisions which differ across jurisdictions (e. g., proportionate liability or joint and several liability, statutes of limitations, allowable organizational forms) may also affect litigation risk. These provisions could, to varying degrees, mitigate or intensify the advantage (to CPAs) of privity relative to foreseeability. Thus, although the result is interesting, it may not capture a true understanding of the legal environment as well as actual litigation data.
The Department of Justice released a series of reports on tort and civil litigation in 45 of the 75 largest counties in the country. Data from that report is compared to the CPAs' perceptions to determine if the perceptions are consistent with the reported litigation experience. Collectively, the CPAs' perceptions about litigation frequency and the damages awarded in their states were inaccurate.
These results suggest that litigation may not be as important an issue to smaller CPA firms as has been assumed. Given their skills and education, it seems plausible that if the CPAs are concerned about litigation that they would be able to education themselves about the issues and make accurate assessments.
Misunderstanding of the litigation risk can have several negative consequences to the CPA. If the CPA underestimates litigation risk, the audit may not be properly designed to reduce this risk to acceptable levels, client screening could be inadequate, malpractice insurance coverage could be insufficient, and/or other dysfunctional actions may be taken. On the other hand, if the CPA overestimates litigation risk, the CPA may overaudit the client's financial statements, engage in excessive client screening, overpurchase malpractice insurance, and/or take other dysfunctional actions.
From society's perspective, overestimating litigation risk is probably preferable to underestimating litigation risk. Deterrence is one of the objectives of a civil liability system (Latham and Linville, 1998). Effectiveness in preventing CPA malfeasance can lead to reduction of audit failures and the massive social costs that can result from the auditors failing to do their jobs. Enron is a contemporary example of the massive costs to the economy that can result from (alleged) CPA malfeasance (Los Angeles Times, January 20, 2002).
Although inaccurately assessing litigation risk can lead to improper actions by the CPA, the results suggest an interesting possibility of oversetting errors. A CPA could reasonably measure litigation risk as an expectation formed with litigation frequency as the probability and damages as the potential litigation cost. The expectation could be correct even though the individual elements are wrong if one element is overestimated and the other element is underestimated by a proportional amount. In California and New York, the CPAs who expressed inaccurate perceptions overstated the probability of lawsuits and understated damages. So even though these CPAs do not understand the elements of litigation risk, they may have "accidentally" estimated overall litigation risk accurately and thus avoid the negative consequences of inaccurately assessing litigation risk.
This study is subject to limitations. The DOJ data did not have a specific classification for audit litigation. The estimated measures of audit litigation are combined with other unspecified business litigation. The unspecified litigation is assumed to be consistent across states and may actually be useful even if it confounds measurement of audit litigation. Any litigation closely enough related to audit litigation that it is classified jointly is likely to signal a litigious environment in which audit litigation is likely to occur even if it has not yet occurred. The mapping of the DOJ data into the Likert scale used to measure perceptions involves estimation. If any of the stated assumptions or estimation techniques are inaccurate, the use of DOJ data in this study could be questionable. In addition, the CPAs involved in the study are from small CPA firms so the results might not generalize to the entire community of CPAs.
This study examines the accuracy of CPAs' perceptions of litigation risk. As an exploratory study into this area, several unanswered questions remain for future research. Do errors in the CPAs' perceptions of litigation risk lead to inappropriate and costly actions? Does heightened attention on litigation issues, such as occurred around the passage of the Private Securities Reform Act of 1995 or around the revelation of a major scandal such as is now occurring with Enron, lead to a better understanding of litigation risk? Does litigation action against an individual CPA lead to more accurate future assessments of litigation risk? Do CPAs who overestimate litigation risk purchase excessive malpractice insurance? Do CPAs who overestimate litigation risk "overscreen" clients? Do CPAs who overestimate litigation risk "overaudit" their clients? Have CPAs who overestimate litigation risk reduced audit services? Do CPAs who underestimate litigation risk have a higher frequency of litigation than those that correctly estimate or overestimate litigation risk?
American Institute of Certified Public Accountants (AICPA) (1995). Implementing a legal liability gap analysis study. New York: AICPA.
Bily v. Arthur Young & Co., 834 P. 2d 745 (California 1992).
Black, S. F., A. B. Weissman, and A. N. Kroll (1996). The Private Securities Litigation Reform Act of 1995: A preliminary analysis. Securities Regulation Law Journal 24(2): 117-142.
Casey, L. L. (1999). Shutting the door to state courts: The Securities Litigation Uniform Standards Act of 1998. Securities Regulation Law Journal 27(2): 141-194.
Dalton, D. R., J. W. Hill, and R. J. Ramsey (1994). The big chill. Journal of Accountancy November: 53-56.
Demery, P. (1996). New privity law in New Jersey. The Practical Accountant 29(1): 12.
Department of Justice (DOJ), Bureau of Justice Statistics (1995a). Tort cases in large counties. Civil Justice Survey of State Courts, 1992. Washington, D. C.
Department of Justice (DOJ), Bureau of Justice Statistics (1995b). Civil jury cases and verdicts in large counties. Civil Justice Survey of State Courts, 1992. Washington, D. C.
Department of Justice (DOJ), Bureau of Justice Statistics (1999). Civil trial cases and verdicts in large counties, 1996. Civil Justice Survey of State Courts, 1996. Washington, D. C.
Department of Justice (DOJ), Bureau of Justice Statistics (2000). Tort trials and verdicts in large counties, 1996. Civil Justice Survey of State Courts, 1996. Washington, D. C.
Kou, J. C. (1998). Closing the loophole in the Private Securities Litigation Reform Act of 1995. New York University Law Review 73(1): 253-292.
Latham, C. K. and M. Linville (1998). A review of the literature in audit litigation. Journal of Accounting Literature 17: 175-213.
Lys, T. and R. L. Watts (1994) Lawsuits against auditors. Journal of Accounting Research 32: 65-93.
McEnroe, J. E. and S. C. Martens (2001). Auditors' and investors' perceptions of the "expectation gap." Accounting Horizons 15(4): 345-358.
Nelson, J., J. Ronen, and L. White (1988). Legal liabilities and the market for audit services. Journal of Accounting, Auditing, and Finance 3: 255-295.
Palmrose, Z. (1997). Audit litigation research: Do the merits matter? An assessment and directions for future research. Journal of Accounting and Public Policy 16: 355-378.
von Brachel, J. (1996). CPAs on Capitol Hill: A behind-the-scenes look at the passage of Securities Litigation Reform. Journal of Accountancy 181(6): 15-18.
Mark Linville, Kansas State University
Table 1: Questions about Litigation Risk 1. As compared to the statutes of other states, the audit malpractice provisions of your state laws are in general: much less slightly less about the hostile less hostile hostile same 1 2 3 4 slightly more much more hostile More hostile hostile 5 6 7 2. As compared to Federal securities law, the audit malpractice provisions of your state laws are in general: much less slightly less about the hostile less hostile hostile same 1 2 3 4 slightly more much more hostile More hostile hostile 5 6 7 3. As compared to other states, for a given audit, the likelihood of a lawsuit alleging audit malpractice is: much less slightly less about the likely less likely likely same 1 2 3 4 slightly more much more likely More likely likely 5 6 7 4. As compared to other states, for comparable situations, the potential damages of lawsuits alleging audit malpractice are in general: much less less slightly less same 1 2 3 4 slightly more More much more 5 6 7 5. As compared to other states, for comparable situations, the likelihood of a state agency filing a lawsuit alleging audit malpractice is: much less less likely slightly about the likely less likely same 1 2 3 4 slightly More likely much more more likely likely 5 6 7 Table 2: Determining the Accurate Range on the Likert Scale of Perceptions Frequency of Litigation State State Median National Accurate Rate Median Rate Perception CA 53 49 4 and 5 NJ 43 49 3 and 4 WI 21 49 2 and 3 NY 46 49 3 and 4 IL 36 49 2 and 3 Damages State National Accurate State Median Median Perception CA 67,834 50,318 5 and 6 NJ 25,725 50,318 2 and 3 WI 25,000 50,318 2 and 3 NY 150,000 50,318 7 IL 38,545 50,318 3 and 4 Table 3: Comparisons of Perceptions of Litigation to Third-Party Liability Standards Average Response from Average P-value of Foreseeable Response from difference Question: (a) State Privity State (b) As compared to the 4.8796 4.0260 0.0000 statutes of other states, the audit malpractice provisions of your state laws are in general: As compared to Federal 4.4000 3.6533 0.0000 securities law, the audit malpractice provisions of your state laws are in general: As compared to other 4.8991 4.2933 0.0000 states, for a given audit, the likelihood of a lawsuit alleging audit malpractice is: As compared to other 5.0734 4.4133 0.0000 states, for comparable situations, the potential damages of lawsuits alleging audit malpractice are in general: As compared to other 4.2925 3.9733 0.0156 states, for comparable situations, the likelihood of a state agency filing a lawsuit alleging audit malpractice is: Sample size for these test ranges from 180 to 185. (a) Each question calls for the CPA to respond to a 7-point Likert scale (see Table 1). (b) One tail p-value from a t-test of means. Table 4: Comparison of Litigation Frequency Before and After Abandonment of the Foreseeable Standard of the Third-Party Liability Resolutions per Resolutions per 100,000 100,000 population population Counties State in 1992 in 1996 (post (foreseeable -foreseeable standard) (a) standard) (b) Alameda CA 1.59 0.68 Contra Costa CA 1.36 0.45 Fresno CA 1.39 1.33 Los Angeles CA 1.58 1.64 Orange CA 2.83 1.25 San Bernadino CA 1.67 0.31 San Francisco CA 2.21 1.77 Santa Clara CA 1.30 1.31 Ventura CA 0.92 2.80 California Average 1.65 1.28 P-value of difference (c) 0.1299 Bergen NJ 2.35 1.18 Essex NJ 0.74 1.06 Middlesex NJ 1.29 2.85 New Jersey Average 1.46 1.70 P-value of difference (c) not significant (d) Grand Average 1.60 1.39 P-value of difference (c) 0.2295 (a) Source of raw data: Tort Cases in Large Counties (Department of Justice 1995a), Appendix Table 1. (b) Source of raw data: Tort Trials and Verdicts in Large Counties (Department of Justice 2000), Appendix A. (c) One tail p-value from a t-test of means. (d) One tail p-value is not calculated since the results are opposite of prediction. Table 5: Comparison of CPAs' Perception of Litigation Frequency to DOJ Reported Litigation Frequency Perceptions Perceptions State Accurate Inaccurate % Accurate too low too high CA 27 50 35.06 * 0 50 NJ 10 10 50.00 0 10 WI 7 5 58.33 0 5 IL 9 25 26.47 * 0 25 NY 0 41 0.00 * 0 41 53 131 28.80 * 0 131 Accurate responses are those that map into Likert scale in the range consistent with the actual DOJ frequencies (see Table 2). Inaccurate responses are those that do not map into Likert scale in the range consistent with the actual DOJ frequencies (see Table 2). Perceptions too low are those that map into the Likert scale in a range below the range consistent with the actual DOJ frequencies (see Table 2). Perceptions too high are those that map into the Likert scale in a range above the range consistent with the actual DOJ frequencies (see Table 2). * Different than 50% at the 0.01 level of statistical significance or better. Table 6: Comparison of CPAs' Perception of Damage Amounts to DOJ Reported Damage Amounts Perceptions Perceptions State Accurate Inaccurate % Accurate too low too high CA 48 29 62.34 * 19 10 NJ 1 19 5.00 * 0 19 WI 1 11 8.33 * 0 11 IL 25 9 73.53 * 1 8 NY 1 40 2.44 * 40 0 76 108 28.80 * 60 48 Accurate responses are those that map into Likert scale in the range consistent with the actual DOJ damages (see Table 2). Inaccurate responses are those that do not map into Likert scale in the range consistent with the actual DOJ damages (see Table 2). Perceptions too low are those that map into the Likert scale in a range below the range consistent with the actual DOJ damages (see Table 2). Perceptions too high are those that map into the Likert scale in a range above the range consistent with the actual DOJ damages (see Table 2). * Different than 50% at the 0.01 level of statistical significance or better.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||MANUSCRIPTS; Certified Public Accountant|
|Publication:||Academy of Accounting and Financial Studies Journal|
|Date:||Sep 1, 2002|
|Previous Article:||Analysis of derivatives' unrealized gains/losses from pre-SFAS 133 disclosures: banking industry.|
|Next Article:||Earnings management: the case of other postemployment benefits and SFAS No. 106.|